[Top] [Prev] [Next] [Bottom]

4. Marketing Functions, Markets, and Food Price Formation

The food marketing sector transforms the raw agricultural commodities produced by farmers into the foods purchased and eaten by consumers. The costs of storage, transportation, and processing-the marketing transformations-are an integral component of food price formation. Because the producer and consumer are typically different individuals even in developing countries, commodities must pass from one owner to another, frequently many times, before reaching the family table. This process of exchange takes place in markets and can be conducted only when a price has been agreed on between buyer and seller. In any but a purely subsistence economy, these three topics are inextricably linked-the productive functions of marketing, the role of the market as an arena of exchange, and the formation of food prices at which exchange takes place. For both socialist and capitalist economies, it is impossible to discuss any one of these without at least implicitly discussing the others as well.

This chapter is designed to provide food policy analysts with an understanding of all three topics. It employs the same general structure that was used to present the analysis of food consumption and production: an introduction to the issues that make marketing, markets, and price formation important to food policy analysts; a guide to the data needed to describe and understand the marketing sector and price formation; an outline of the theoretical perspectives and empirical techniques of analysis that permit analysts to reach rough judgments on the extent and social costs of inefficiencies in marketing and distortions in price formation; and a summary of the instruments available for government intervention in both marketing and price formation and a policy discussion of the likely costs and benefits of such interventions.

For the sake of convenience, the term "marketing" will be used inclusively to refer to all three topics treated in this chapter, and much of the discussion will be oriented toward those economies that use markets to determine prices and allocate resources. The term "marketing functions" is used to refer specifically to the commodity transformations in time, space, and form that are associated with storage, transportation, and processing. These functions must be performed in socialist and capitalist economies alike. Even in socialist economies, prices provide producers and consumers with signals about how to allocate their inputs and household budgets. Consequently, how socialist economies set their prices and what impacr they have are important analytical questions. Because marketing encompasses so many activities that are at the core of all food systems, understanding the full range of marketing issues is a central task for food policy analysis.

Marketing Issues

Just as with food production, marketing is a means to an end. The objectives a society can reasonably hold for its marketing sector are analogous to the four basic objectives for the food system as a whole: efficient economic growth, a more equal distribution of incomes, nutritional well-being, and food security. Because it links the production and consumption sectors, marketing can contribute to all four objectives rhrough the efficiency with which it communicates signals of scarcity and abundance to decisionmakers. Because it is the source of productive activities involving many jobs, marketing can contribute directly to economic growth, income distribution, and nutritional objectives. Because of its capacity to link domestic markets to international markets and to provide signals to policymakers concerning food shortages, the marketing sector is integral to the design of mechanisms to improve food security.

Why is it, then, that marketing activities are frequently thought to be unnecessary and against the interests of society, especially the interests of poor people in developing countries? The labels "middleman" and "speculator" almost universally carry negative connotations or actual opprobrium. The tendency is too widespread to be attributed wholly to a bad press. In fact, markets do not always function in the best interests of a broad cross section of society, especially in poor countries where communications and transportation facilities are poor, markets are highly segmented, and access for marketing participants is greatly restricted, sometimes to particular ethnic groups. Highly unequal financial bargaining power is often brought to the exchange relationship between seller and buyer. In short, the efficiency and economic gains potentially available from successful market coordination of a society's food system are an empirical issue, not a matter of faith and logic.

Because the public image of marketing is so negative, especially in most developing countries, it is important for the food policy analyst to determine how effectively marketing institutions and marketing agents are performing their dual roles of transforming commodities in time, space, and form while reflecting relative abundance and scarcity through the price signals communicated to producers and consumers. These price signals can be generated in the process of exchange in markets, in which case the competitiveness and efficiency of the markets must be examined. Alternatively, if governments set these prices to reflect other criteria and objectives, their effectiveness and costs in other dimensions must be examined.

Elements of Market Efficiency and Market Failure

Markets are the arena for two important tasks required in all societies: the physical marketing functions, and the communication of signals to producers and consumers about the costs of buying something or the benefits of selling it. Governments concerned about equal distribution of economic welfare to all citizens are understandably loathe to allow those price signals to be generated by anonymous market forces when the values of such important commodities as food and fuel, for example, or services, especially labor, are at stake.

A shortage of food means high prices in a market economy, with only the well-to-do able to purchase it. A food shortage in a socialist economy means rationing, with perhaps long lines and little choice about what goes into the food basket. In the short run, the socialist approach may deal more effectively with hunger, but hunger is also a long-run problem of development and efficient use of resources. Here the allocative role of prices becomes important, in addition to their role in income distribution. Much of the chapter is devoted to understanding the tension between these two roles. The last section seeks ways to improve the efficiency of allocation by using appropriate price signals without relinquishing all food distribution authority to the free workings of the market.

Competition and the Number of Market Participants

Virtually all of the positive welfare implications of market coordination are derived from economists' models that use "perfect competition" assumptions to drive the logic of decisionmaking behavior on the part of market participants. Competition is a powerful force in economies. It is the "invisible hand" that guides private self-interest into maximizing social welfare. For competition to play this powerful role, however, there must be an adequate number of participants on both sides of the exchange relationship so that no single agent can significantly influence the outcome of the exchange. The only time the condition is clearly violated is when only a single participant sits on one side or the other. Even two sellers can provide strong competition for each other if they compete. Alternatively, twenty sellers may not be competitive if a mutual understanding exists about their appropriate market behavior. Very large numbers of participants-millions of farmers and millions of consumers-guarantee competitiveness at each end of the food system. The issue is the number of participants in the chain in between and the potential access of additional participants if the returns to providing marketing services rise above the level dictated by competitive equilibrium.

The farmer is concerned to get the highest possible price for the output to be sold (or the lowest possible price for the inputs to be purchased) for a given level of ancillary conditions, such as credit, speed of payment, and discounts for moisture. The farmer must determine which marketing agent to sell to. The more agents there are competing to buy the farmer's grain, the better the information available to the farmer about the prevailing price and the easier it is to switch from one buyer to another whose terms are relatively better. Where there are many potential secondary buyers, such as rice mills or export firms, many farm-level buying agents will offer the farmer great freedom of choice. If only a single end buyer exists, as in the case of pineapples for canning, the farmer may have little choice but to sell to the buyer from the cannery or to take a very large discount in a small local fruit market. Identifying the farmer's range of choice at the initial point of sale is the first step in understanding how competitive price formation is likely to be.

A similar approach holds at the opposite end of the marketing chain, where consumers buy foods for home consumption. Individual consumers have no influence over the prices they pay, but if many alternative retail stalls offer similar commodities and services, the freedom of consumers to choose one retailer over another prevents excess profits from high margins accruing to the retail marketing agents. If only a single retailer is available for miles around, the potential for high profits is great. In such circumstances the analyst must wonder why other retailers do not join the action. If significant barriers to entry impede the addition of more retailers, government steps to improve access to retail trade may pay considerable dividends to food consumers. If the monopoly position is caused by government regulations or is held by the government itself, the impact on consumers through reduced freedom of choice and higher prices must be balanced against any potential benefits that result from the government's role.

Transfer of Ownership and the Balance of Market Power

When both seller and buyer agree on the terms of a sale, a price is established for the exchange. Both parties must be satisfied, but who actually decides what the price will be? Economists have been puzzled by this question for centuries. Some have invented quite fanciful arrangements, such as Leon Walras's auctioneer who permitted recontracting among parties until overall equilibrium was reached. Kenneth Arrow's suggestion is more relevant to the issues here. He sees all exchange relationships as having at least small elements of monopoly, or market, power on one side of the exchange or the other. Each exchange is to some extent unique because each party has barriers of time and distance between an alternative exchange party. Each party comes to the exchange with different knowledge about the characteristics of the underlying market forces for the item of exchange. Arrow argues that the party with the relatively greater knowledge actually sets the initial price. The other party then decides whether to accept or reject the offered or posted price. If little competition exists, there will be little pressure to set the posted price close to the actual costs of offering the product in that time, place, and form. Heavy competition, however, improves the other party's knowledge of market conditions, and it forces an adjustment in posted price by either direct negotiation or the patronizing of alternative dealers.

In such a framework of price formation, market knowledge is market power. One of the most important steps governments can take to improve the fairness of market price formation so that it discriminates less against the small farmer at one end and the consumer at the other is to provide these individuals with timely and accurate information about actual market conditions. Such information enables them to bargain more equally with purchasing agents or retailers who are naturally intent on widening their margins whenever p05sible. More equal balance of knowledge provides a more equal distribution of the gains from efficient market price formation.

Efficiency of Price Formation

Prices are formed efficiently when large numbers of buyers and sellers, all with similar access to relevant market information, interact to agree on a basis of exchange, a price. This price sends signals to consumers about the resource costs of supplying the commodity to them. It simultaneously sends signals to producers about the willingness of consumers to pay the resource costs of production. Efficient price formation is essential to the efficient allocation of resources in a market-directed economy.

This picture of price formation is essentially static, or at most it captures a sequence of static equilibria. However, expectations about future conditions are also likely to be important in the actual formation of prices. If expectations are precisely fulfilled in each period, a perfectly predictable dynamic pattern of prices results. Of course, the real world is never so accommodating. The essence of the interaction between expectations and price formation is that some market participants' expectations about the future are constantly being contradicted as new market information becomes available. In what amounts to a continuing wager against alternative expectations, large sums of money are made and lost when drought hits or bumper crops roll in.

Is this somehow inefficient? It is tempting to think that governments should simply set prices for basic grains (and other important commodities) at some "fair" level and prohibit the kind of trading that leads to speculative gains and losses. The capital tied up in such trading could be put to more productive social use through investments in factories or dams. Alas, this approach fails to recognize the dual role of dynamic price formation. It integrates information about future crops and alternative supplies, demand pressures, and storage costs to allocate the supplies in hand to future time periods. At the same time, the temporal pattern of prices established, or the price expectations formed, signal producers, consumers, and the suppliers of storage as to the opportunity costs of their production, consumption, and storage decisions.

Failure to receive accurate signals about these opportunity costs can cause enormous misallocation of resources in food production and consumption and very serious disruptions to the smooth temporal flow of food supplies to consumers. Many socialist economies, for example, try to use markets as a vehicle for achieving short-run efficiency in the distribution of goods in the economy while the government determines price signals that the markets will reflect to producers and consumers. Such government-set price signals in command economies tend to communicate in only one direction, from the top down. When the price signals fail to reflect the opportunity costs to the society of actual production and consumption decisions, no mechanism communicates this information back up to the price control commission. As the signals become more and more unbalanced, they frequently lead to severe shortages in the food system, as in Poland, or to massive budget subsidies for food producers or consumers, as in Egypt and China. Truncating the dynamic aspects of price formation caused by expectations also hobbles the mechanisms that produce static efficiency in the allocation of resources. It is not possible to have one without the other.

Market Failure

For all of its efficiency in allocating economic resources, a competitive market economy cannot accomplish some important social goals without careful government intervention. Some of these broader goals also relate to economic efficiency; others concern the distribution of income, nutritional well-being, and security of the society and its food supplies.

Even competitive markets with efficient price formation fail to provide a socially efficient allocation of resources if externalities exist within the economic system. Unfortunately, the food system is full of externalities. Irrigation decisions upstream affect water supplies downstream. Privately profitable applications of pesticides have consequences for public health and for the environment. When farmers cultivate hillsides and marginal lands in pursuit of food for their families (for example, on Java) or profits (for example, in Nebraska), soils erode. Major changes in food policy by large countries in the world food system, such as the U.S.S.R., the United States, and China, have an impact on producers and consumers in other countries.

The existence of such externalities and the failure of market-determined outcomes to provide efficient solutions are reasons for the analysis and design of food policies where government is an important participant in the food system. At the same time, an understanding of these factors should provide insight into the areas of economic resource allocation in which markets do an efficient job and government intervention is likely to make matters worse, not better. Rather than telling farmers how much pesticide to use and trying to enforce the rule, pesticide prices might be set to reflect full social costs. Incentives to plant crops that cause less soil erosion or replanting schemes to stabilize barren lands may be more effective than police action to prevent farmers from cultivating hillsides or villagers from poaching firewood from public lands.

The tension between intervening in markets and letting markets work is very strong in most countries. Intervention is frequently predicated on the existence of market externalities. In the context of a careful empirical understanding of the quantitative impact of such externalities, their existence does indeed call for specific government intervention. Typically, however, such interventions call for a scalpel rather than a sword.

Private market economies also fail to provide adequate quantities of public goods, such as national defense, police protection, scientific research, even roads and education. These "goods" all yield benefits to the population at large that cannot be priced and directly charged to the users by the (private) suppliers. An important role for government is to use general tax revenues to provide such public goods in socially optimal amounts.

Certain components of the food system are like public goods. Agricultural research is no doubt the most important, and governments of virtually all countries accept responsibility for financing agricultural research and adapting it to local environments. Rural education, roads, and communications networks all have at least an element of public good in their supply and demand and hence call for a government program to supplement private supplies.

Many observers regard the entire food marketing system as a public good because of the synergies and interdependence among its various components. A smoothly functioning marketing system depends on the simultaneous availability and interaction of these components: efficient communications, transportation, and storage facilities; common grades and standards to facilitate trading at a distance; legal codes to enforce contracts; credit availability to finance short-run inventories and processing operations; and a market information system to keep all market participants, from farmers to consumers, fairly and accurately informed about market trends. No private investor could hope to capture the total gains from the interactive synergy of this system, and individual investments in pieces of the system do not produce the overall synergy. Consequently, a vision of the ultimate productivity of an efficient marketing system provides a powerful impetus for extensive government involvement in the design, construction, and possibly the operation of the food marketing system.

Such a vision has much to recommend it. But it is not a vision that is easily put in focus by looking at the food marketing systems of the United States or Western Europe. The efficient functioning of marketing systems is particularly sensitive to local, cultural, and social conditions and especially to the local availability of resources. These resources include labor and capital, of course, but for marketing systems they also include managerial, administrative, and entrepreneurial resources, which are in very short supply for most governments. Consequently, massive government efforts to "modernize" food marketing sectors, especially if private investment and participation are not included, are likely to run afoul of the very complexity of the system itself.

It is well understood by both economists and politicians that the efficient outcome of market forces does not necessarily imply a satisfactory distribution of income or of food consumption. Most economists would like to "fix" the income distribution with some sort of neutral income transfer via the government budget, rather than alter prices for important commodities that influence the distribution of economic welfare. This approach preserves the efficiency of the market solution by not distorting producer or consumer choices, but most government policymakers find it impossible in practice. A typical response has been to use government interventions not to transfer income directly to poor people, but to alter important prices that significantly affect real incomes, because governments have more short-run control over prices than over individual incomes.

In the name of improving income distribution and the adequacy of food intake, therefore, many governments try to keep food prices low, wage rates high, interest rates low, and imports cheap through an overvalued foreign exchange rate. All these prices do have important implications for the real income of virtually everyone in a society, but they also are absolutely critical as signals for the efficient allocation of resources. Here again is the dilemma over the short-run welfare of the population, especially the poor whose food consumption can least stand to be reduced, and the longer-run efficiency of resource allocation that comes from allowing the scarcity of commodities and factors of production to be reflected in the prices paid for their consumption or use. This dilemma appears repeatedly in this book. The elements of a resolution are assembled in the consumption and production chapters, while the vehicles for their implementation are analyzed later in this chapter. First, the task of analyzing marketing functions, markets, and price formation lies ahead.

Marketing Functions and Price Formation

Suppose the farmer represented in the tableau of chapter 3 chose the optimal technique and has just harvested 6 tons of rice in Season 1. Three questions are now pressing: how much rice should be sold (and how much kept for home consumption), what price will it bring, and who will buy it? These questions lead straight to the heart of the marketing issues, for answering them begins the process of understanding how rice harvested in the farmer's field is transformed into a meal of rice to be consumed in another time and place.

The farm family is both a producer and a consumer. How much rice will this farm household wish to consume? In terms of consumption, the most important determinants of the answer will be habits, income, and rice prices relative to other staple food prices. But when marketing considerations are added, the issue becomes complicated because the farm household income depends on the price of rice. In some circumstances a higher price for rice could lead the farmer to consume greater amounts because of a strong income effect. Most empirical evidence suggests, however, that farm families have negatively sloped demand curves for the foodstuffs they grow. Although perhaps less elastic in response to price changes than those of nonfarmers living in rural areas where a wide choice of foodstuffs is available, the demand curves of most farmers slope downward like the one shown in figure 4-1. Both the vertical postharvest supply function and the demand function are depicted for a representative farmer. Naturally, the supply function for future periods would be expected to have a positive slope, reflecting the farmer's ability and willingness to expand output in the face of greater price incentives.

With the farmer's rice supply fixed in the short run by the size of the harvest and with the quantity to be consumed at home determined by the market price for rice, the amount the farmer is willing to sell to the market will also be a function of that market price. At high rice prices the farmer prefers to consume somewhat less rice (and more corn, wheat, cassava, or other staples that are relatively cheaper when rice prices are high) and consequently is able to sell more rice. The amount offered to the market, the difference between the fixed supply Qh and household consumption, is a rising function of price (at least up to the total available supply Qh) and is shown in figure 4-1 by the dashed excess supply curve. This line reflects the amount of rice the farmer will offer to the market at each market price and is constructed by subtracting household rice consumption, itself a negative function of the market rice price, from the farmer's rice production. The excess supply curve relates the farmer's market sales of rice to the price received. If the market price is Pm, the farmer sells Qs and consumes Qc. The geometry of the excess supply curve guarantees that Qs plus Qc equals Qh.

From whom does the farmer receive this "market price"? Although economists are prone to think of exchange and price formation as neutral concepts which happen automatically, market participants know better. Marketing agents are real people making decisions designed to improve their economic well-being (and contributing to economic output in the process). The farmer must find someone to buy the rice, or some marketing agent must find the farmer.

What motivates such a marketing agent? By purchasing an amount Qs of the farmer's rough rice in the farmer's field at an agreed market price Pm the agent hopes to do one or more of three things. First, the agent might transport the rough rice to a bulking point or a rice mill and resell it

Figure 4-1. Short-run Supply and Demand Curves for a Representative Rice Farmer

immediately, hoping to recover in the resale price a margin for the effort and the risk involved in the transaction, for the rice mill may not be willing to pay more than the agent paid the farmer.

Second, the agent might have a rice mill nearby. When the rice is milled, it becomes more valuable to consumers, who prefer to eat milled rice instead of paddy or brown rice. If the willingness of consumers to pay more for milled rice is sufficiently great, the marketing agent might be able to recover the costs incurred for the rough rice when it was purchased from the farmer, plus an adequate margin to pay the full costs of processing, including a return for investment costs and risks.

Third, the agent might store the rice in a warehouse and hope to sell it for a higher price later in the year, after the abundant harvest supplies have been absorbed. This process involves even more risk, for the agent must pay additional costs to store the rice~interest on the money invested in the rice, handling the rice as it is moved in and out of the warehouse, rental costs (or maintenance and depreciation charges) for the warehouse, losses while the rice is in storage, and insurance against fire and other hazards. When the rice is removed from storage, there is no guarantee that the agent will be able to sell it for a price sufficiently higher than the price paid to cover these costs plus a return for the time, effort, and risk involved. If the price is too low to cover these costs fully, the agent is likely to stop buying rice for storage or to offer the farmer a lower price at the next harvest. Although each of these functions-transportation, processing, and storage-was treated independently, the same agent could be involved in any or all of them.

Costs and Prices

It is apparent that an array of prices for rice is being formed during the performance of the marketing functions. Consumers express a willingness to buy milled rice via their demand curves, farmers express a willingness to sell rough rice through their excess supply curves, and marketing agents connect the two parties by being willing, for an economic return to their time, capital, and risk-taking, to pay the farm price and to transport, store, and process the rice in order to sell it to the consumer in the time, place, and form desired. Prices are determined at each stage of this complicated marketing process. Although it is possible to get lost in the complexity, it is important to understand the basic forces that explain this process of price determination in markets. The example presented here focuses on temporal price formation, that is, the relationship between prices during the harvest and prices during the "short" season before new supplies from the next harvest become available. The transportation and processing functions can also be understood within this framework. They are held in abeyance for the moment in order to concentrate on the essence of price formation and its effects on quantities supplied and demanded in each period.

Figure 4-2 carries forward by two steps the analysis of the supply, demand, and market offers of the representative farmer shown in figure 4-1. First, figure 4-2 deals with market aggregates rather than an individual farmer. Consequently, the farm market supply function (offer curve) during the harvest period shown on the right side of figure 4-2 is the sum of all farmers' offer curves at each possible harvest period price Ph It is positively sloped in the short run because farmers consume less rice at higher prices. In the longer run it will be even more elastically sloped because farmers will also react to higher prices by increasing output. The market demand curve for rice during the harvest period, also shown on the right side of figure 4-2, reflects the willingness of consumers to buy rice from the market at various prices (it is net of farm consumption from retained supplies). If this were all

Figure 4-2. A Back -to-back Supply and Demand Framework for Understanding Seasonal Price Formation

there was to the story, competitive forces in the market would be expected to bring about an equilibrium where both consumers and farmers were satisfied with the available price-quantity relationship. In this example, the supply and demand curves intersect at Ph,1 and Qh,e.

But there is more to the story. The left side of figure 4-2 introduces the second step into the analysis by showing that consumers wish to eat rice even during the time it is not being harvested. By arbitrarily dividing the rice calendar into two periods, a harvest and a nonharvest period, figure 4-2 is able to show the supply, demand, and price consequences of linking the two periods through the storage function. The vertical axis measures the price of rice for both periods, for the harvest period extending to the right and for the nonharvest period extending to the left. The right side of the horizontal axis measures quantities in the usual way of a Supply and demand diagram; the quantity increases from the origin to the right. To the left, however, quantities are measured by reflection, as in a mirror image. Increasing quantities are shown by moving to the left of the origin, and so the nonharvest period demand curve must slope down to the left, as a mirror image of the demand curve during the harvest period. Because the two supply and demand diagrams in price-quantity dimensions are placed next to each other with a common price axis, such figures are called back-to-back diagrams. Their usefulness will become apparent when the opportunity to save rice from the harvest period for consumption in the nonharvest period is considered.

No rice is produced in the nonharvest period~n the left side of figure 4-2. At all prices (within a normal range) during the nonharvest period, zero supplies from production are forthcoming. But with zero supplies and substantial demand for rice, prices in the nonharvest period would skyrocket. If prices in the nonharvest period are sufficiently higher than in the harvest period, someone, a marketing agent, is likely to take the risk of purchasing rice during the harvest and storing it for later sale at the higher price in the nonharvest period.

Figure 4-2 shows how this process works. In the figure, 5 equals the costs per kilogram that a marketing agent will incur to store rice from the harvest to the nonharvest period. These storage costs include a suitable reward for time, effort, opportunity costs of capital invested, and return for risk-taking. If costs equal to S are actually covered by the difference in price between the two periods, the marketing agent will be willing to continue these activities year after year. If the price difference is less than S, some marketing agents will find it unprofitable to continue in business as transfer agents between the harvest and nonharvest periods. If the price difference is greater than S, the agents will expand their storage operations, and others may join the activity. Figure 4-2 is constructed on the assumption that the price rise will just equal the storage costs S.

It is now possible to see where supplies will come from in the nonharvest period despite the total lack of production. Speculator~the name for marketing agents who buy when the price is low in the hope that it will rise after the harvest-will be willing to supply from storage an amount of rice that depends on the relative market prices prevailing in the two periods. The two prices must be different by the cost of storing rice from one period to the next. In figure 4-2 the nonharvest price Pnh,l is higher than the equivalent harvest price Ph,l by the storage costs S.

Ph,l is a particularly important price. If it were to prevail during the harvest period, as indicated above, harvest period market supplies would equal harvest period market demand, and no rice would be available for purchase by speculators for storage. In the absence of storage costs, the excess market supply curve for the nonharvest period would start at zero quantities for price Ph,l in the harvest period. Consequently, no rice would be available for consumption in the nonharvest period, and consumers then would have to find other sources of food or go hungry. If speculators bid up the harvest period price to a level higher than Ph,l, however, consumer demand in that period would be less, market supplies forthcoming from farmers would be greater, and speculators would move the excess supplies into storage for later sale and consumption in the nonharvest period. The amount of these excess supplies available for the second period is a function of the price prevailing in the harvest period.

The excess market supply function from harvest period to nonharvest period shown in figure 4-2 reflects this price-quantity relationship. The solid line shows the excess supplies available in the nonharvest period before storage costs are incurred. The dashed line shows the supplies available in the non-harvest period after the costs of storage S have been paid. For each pricequantity relationship showing excess supply at a price during the harvest period, the dashed line shows the equivalent price that must prevail in the nonharvest period to cover the costs of storage.

What happens? This equilibrium price-quantity relationship, shown in figure 4-2 at Pnh,2 and Qnh,c, can be located by determining the price at which the excess supply available in the nonharvest period is equal to the demand during the period. Then, because the harvest price must be lower than the nonharvest price by the costs of storage, the harvest price is located at Ph,2 From this price the decisions of consumers and farmers in the harvest period can be determined. In the example shown, market demand will be Qh,c and market supplies will be Qh,s. The difference between the amounts supplied and demanded is Qh,s minus Qh,c which must equal the amounts stored and consumed in the nonharvest period, Qnh,c. The excess supply curve is constructed so that this is precisely what happens. The construction is easy with the linear supply and demand curves in the example. It can be quite complicated with more complex shapes.

The Market Perspective

Figure 4-2 shows that marketing functions and price formation are simultaneously connected. One affects the other. If storage costs rise because, for example, interest rates rise, the price of rice will be higher in the nonharvest period than before and lower in the harvest period. These price changes will affect total supplies available and the allocation of those supplies between consumers in the two periods. To continue the example, the higher storage costs would cause marketed supplies in the harvest period to fall, as farmers faced lower prices. Harvest period consumers would buy more because they also face lower prices. The entire burden of reduced supplies would fall on nonharvest consumers through the higher prices in the second period. Chapter 2 noted the potential nutritional significance of reduced consumption during nonharvest periods of high seasonal prices. Figure 4-2 shows why high seasonal prices are an essential element in supplying food to the short season in the first place. In the absence of subsidies to reduce the real economic costs of storage, the seasonal price rise and reduced seasonal consumption are necessary to have any food at all in the nonharvest period.

Marketing functions can thus be seen as the essential link between producers and consumers in two very different and yet simultaneous and connected ways. First, the marketing agents link producers and consumers physically, by actually buying, storing, transporting, processing, and selling commodities. Those societies that lack adequate numbers of such marketing agents know best their value. Empty stores and markets, long queues, furtive dealings on side streets or in back rooms testify to the valuable equilibrating role provided by marketing agents as they balance their purchase, transformation, and sales decisions against the likely economic reward. Simultaneously, however, because exchange of commodities is taking place, open or implicit price signals are being generated and transmitted to the active economic agents in the food system, influencing their production and consumption decisions. And by the cumulative reverberations of those decisions back on the marketing agents, and back on price formation, and back on decisionmaking, and so on in the endless repetitive flows characteristic of economies with markets, a dynamic equilibrium process is established by which resources are allocated both to goods the consumers most want and to areas where the resources are most productive in the farm system that supplies the goods.

Because socialist economies tend to use marketing systems in one direction only, the information generated by imbalances in the exchange markets is largely lost as signals to producers and consumers below and to policymakers above. To be sure, parallel markets in rural areas, or even gray-to-black markets in urban areas, transmit some of this information on relative scarcity. Such informal markets provide much of the food and rural income in a few economies where official prices and rigid market regulations have driven supplies underground. The point is not the failure of state interventions in marketing and price formation, although it happens often enough, but rather the importance of the information generated in markets for the efficient use of a society's economic and human resources.

Understanding the Marketing System

When markets work, the automatic adjustment processes perform an awesome task of coordination with a minimum of fuss, and economic resources are allocated efficiently. When markets fail, participants with inside information and economic power are able to exploit both producers and consumers, to the special detriment of the poor at each end. The task of this chapter is to help analysts know when markets are working, identify markets that are failing and understand why, and determine which government interventions would improve both the efficiency of market operations and the distribution of gains generated when markets work efficiently.

The questions of most importance in the marketing sector-the costs and efficiency of providing marketing services and the dynamic capacity of the system to create and transmit signals about incentives to producers and consumers that are consistent with resource availabilities and long-run structural transformation-are not easy to answer. Because direct approaches are frequently constrained by unavailable or unreliable data, indirect approaches that rely on normative competitive models are often used to provide additional insight. In such circumstances, no direct listing of the data needed can lead straight to the analytical techniques. A constant interplay is necessary between the availability and reliability of data on the one hand and the analytical approach used to address important marketing questions on the other.

A number of empirical questions are addressed in analyzing marketing systems. What are the marketing channels for important commodities and who are the participants? What are the costs, margins, and profits that result from this process? What do price data at various levels of the marketing system reveal about the process of price formation and the degree of connection among markets? How are international market prices determined and how do they influence domestic price formation?

As with both consumption and production data, published sources provide the analyst with an efficient first cut at the task of understanding how the marketing system works. But data from field observations, even informal weekend surveys, add critical flavor and insight into the mechanisms that generate published statistics. For example, riding with a bag of rice as it changes hands repeatedly between the farm gate and retail stall and understanding all the decisions made along the way will guarantee the analyst a better perspective on market decisionmaking and price formation than a host of statistical analyses done in the office.

Elements of a Competitive Market

One indirect way to measure market efficiency is to ask whether the elements of a competitive market are present in the marketing system under study. These elements include the following set of conditions: items of the traded commodity are fungible (interchangeable) and divisible; buyers and sellers act in an economically rational fashion (they want more, not less, incomes and goods); firms are small and numerous enough that their decisions have no impact on prices; all participants have equal access to activities of the market on the same terms; and everyone has complete knowledge of forces likely to influence supply and demand.

If these five conditions-divisibility, rationality, small firms, equal access, and complete knowledge-characterize a marketing system, the market will perform efficiently with no scope for excess proftts. Fulfillment of these conditions is sufficient, but not necessary, for a market to be competitive. For example, price formation can occur efficiently in a market where only three or four large firms are selling or buying if these firms compete rather than collude with one another. The main difficulty with this indirect approach, therefore, is how to determine from a survey of market participants whether the system is operating under circumstances that approximate the competitive ideal closely enough to rule out collusion and excess profits.

Of the five conditions needed for a competitive market, two are usually not at issue in food crop marketing. Divisibility is a characteristic of nearly all food commodities, and virtually all market participants react appropriately to economic signals. The third condition, numerous small firms, is also a feature of most marketing systems in developing countries unless the government has created an effective parastatal monopoly or has policies restricting the access of new entrants to the marketing system. Large-scale export marketing firms are sometimes an important exception.

The principal focus is on the conditions of market access and knowledge. Because no system can have perfectly equal access and complete information, the matter is one of degree, that is, whether entry is free enough and information is good enough for the market to work with a reasonable degree of competitiveness-enough to bring about an efficient result. Much can be learned from surveys that inquire about modes of entry into trade (periods of apprenticeship, capital requirements, and age distributions of merchants) and about market information (credit arrangements and risk sharing, bargaining positions of farmers at the first point of sale, and sources of price information for nearby and distant wholesale markets).

Ease of entry into the marketing system (as a petty trader or operator of a small-scale rice mill, for example), in combination with reasonably accessible market information, carries a strong presumption of a competitively efficient marketing system. If the marketing system is characterized by limited access and information, government efforts to provide wider access to working capital, better gathering and dissemination of price statistics and information about crop conditions, or the dissolution of state-sanctioned market-area monopolies may well improve market efficiency.

To find out how many traders are operating in a marketing system, and at what points a commodity changes hands, it is helpful to sketch its flow through the marketing chain. The competitiveness of a market and the structure of the marketing chain are obviously related. If at some point in the chain only a single buyer or seller exists, then noncompetitive behavior is likely. Alternatively, the presence of many active buyers and sellers all along the chain carries a strong presumption of competitive behavior and efficient market performance.

Market Flows and Participants: Marketing Chains

Construction of food crop marketing chains helps organize the links between production and consumption. Some typical marketing chains for a commodity are shown in the following tabulation:

5. 1. Farmer S,P,T > rural consumer

6. 2. Farmer S,P,T > rural retailer T> rural consumer

7. 3. Farmer S,T > resident processor or assembler P,S,T > rural retailer T> rural consumer

8. 4. Farmer S,T > resident processor or retailer P,S,T > nonresident wholesaler S,T > urban consumer

9. 5. Farmer S,T > nonresident wholesaler P,S,T > urban wholesaler or consumer

where T = transfer operation, such as transportation or exchange of ownership; P = processing activity; and S = storage function.

Estimating volumes and percentages of commodity transformations at each link in the chain provides an overview of the marketing system's structure. How much of total production do farmers sell, and how much is sold directly to rural consumers (marketing chain 1)? How much is sold to rural retailers in a nearby marketplace (chain 2)? How much is sold to resident processors or assemblers, who then sell either to rural retailers in the region (chain 3) or to nonresident wholesalers for shipment to urban markets (chain 4)? How much is sold to nonresident wholesalers who travel to producing areas to buy supplies for urban markets (chain 5)?

Using formal or informal market surveys, the analyst attempts to replace the arrows in the marketing chains with quantities or percentages, as for the rice marketing chain shown in figure 4-3. In this example, farmers sold 50 percent of their marketings to local assemblers, 40 percent to local processors, and 5 percent each to district assemblers and processors. These quantities then moved on through the system until 85 percent was exported from the district and 15 percent was consumed locally.

Marketing chain diagrams thus specify market linkages that connect one price series to another to determine, for instance, whether rural food prices are influenced by urban demand or whether expectations of high preharvest prices work back to influence the harvest price. Knowing when and where the crops are sold, their transportation and storage destinations, and who the ultimate consumer is permits the analyst to specify the likely causal direction of market connections.

Marketing Costs and Margins

Large marketing margins-the spread between farm prices and consumer prices-can occur for two reasons: either high real marketing costs cause consumer prices to be much higher than farm prices or monopolistic elements in the marketing system are earning excess profits. Both a direct and an indirect approach can be used to determine whether there are excess profits and serious inefficiency in food crop marketing or whether wide margins are

Figure 4-3. Marketing Chain for Rice, Atebubu District, Ghana

Note: Amounts represent percentage of total marketings.
Source: V. Roy Southworth, "Food Crop Marketing in Atebubu District, Ghana," Ph. D. dissertation, Stanford University, 1981.

due to high real costs which might be reduced through appropriate government investment in marketing infrastructure. Because of serious data restrictions in most circumstances, both approaches~ne that looks at costs, the other at prices and price margins-are usually needed to allow cross-checking with each other.


The direct approach looks at the three marketing functions whose combined costs constitute the marketing margin. This kind of efficiency analysis of marketing has clear analogues with that of any productive activity, including agricultural production. Such empirical applications, however, are usually difficult and sometimes Out of reach because data requirements are very demanding.

Records are needed of representative costs and returns from the main participants in transportation (merchants, transporters, and brokers), processing, and storage. Analysts estimate the costs of all inputs, including management costs. Subtracting costs from returns then gives profits at each level of the system. An example of a marketing margin for rice in Ghana is presented in table 4-1. In the example, farmers received 70 percent of the urban retail price, and net returns to assemblers, processors, and retailers each ranged between 4.6 and 6.5 percent of the urban retail price.
Table 12: Average Marketing Margins for Rice Produced in Atebubu District and Retailed in Kumasi, Ghana, January through July 1977
Item Cedis per tona Percentage of retail price
Atebubu market wholesale priceb 2,216.93 83.0
Producer pricec 1,894.38 70.9
Assemblers' gross margin 322.55 12.1
Commissions (20.06) (0.8)
Transport (90.28) (3.4)
Handling and storage (8.00) (0.3)
District tax (30.10) (1.1)
Assemblers' net margin 174.11 6.5
Kumasi wholesale priceb 2,525.00 94.5
Atebubu market wholesale price 2.216.93 83.0
Processors' gross margin 308.07 11.5
Transport (50.06) (1.9)
Parboiling (40.13) (1.5)
Milling (60.19) (2.2)
Handling (16.03) (0.6)
Commissions (20.00) (0.7)
Processors' net margin 121.66 4.6
Kumasi retail priceb 2,671.80 100.0
Kumasi wholesale price 2.525.00 94.5
Retailers' gross margin 146.80 5.5
Handling (6.50) (0.2)
Stall rental (7.00) (0.3)
Retailers net margin 133.30 5.0

a. Paddy prices at the producer and Atebubu market wholesale level were converted to their milled equivalent at a 0.62 milling ratio. One cedi = 0.87 U.S. dollars.
b. Market prices are the average of monthly price figures collected by the Ministry of Agriculture from January through July 1977
c. The producer prices are the average of prices recorded in the farm survey for producers in the Kwame Danso area.
Source: V. Roy Southworth, "Food Crop Marketing in Atebubu District, Ghana," Ph. D. dissertation, Stanford University, 1981.

Because normal profit is the return to capital (including working capital as well as equipment and buildings), data are needed on capital used in the marketing enterprises. Rates of profit are calculated by comparing estimated profit levels with the amount of capital in use. To determine if the return to capital represents normal profits, the analyst must compare it with the prevailing interest rates in the credit markets to which the food crop merchants have access. If the prevailing interest rate (which contains a premium for the riskiness of the marketing investment) is less than the earned rate of return, the earned level of profit is above normal.

Although this type of analysis is time-consuming and the data requirements are extensive, it is worth attempting for policy purposes if there is some prior evidence of excess profits or if policymakers believe that marketing inefficiencies are so pervasive that only solid empirical evidence will prevent implementation of marketing policies based on that belief.


Price analysis is an indirect approach for determining market efficiency. Efficient marketing systems are characterized by a high degree of price integration~losely correlated movements of connected series of price~ over space, form, and time. In an efficient market economy, price integration is caused by arbitrage. In these economies market participants respond when they notice that prices in two markets are sufficiently different that profits can be made by buying in the low-price market and selling in the high-price market. If competitive conditions prevail and enough merchants respond in this way, the abnormal price difference disappears because supplies in the low-price market decline, placing upward pressure on prices, and supplies in the high-price market increase, causing prices to fall. Hence, prices in all efficient markets are linked by the arbitraging decisions of merchants, and price differences should reflect only normal costs. However, expectations about future price levels are an important ingredient in price formation. Because future price levels are uncertain, merchants bear risks when holding commodities, and the costs of risk-bearing are also included in marketing margins.

Price analysis of marketing margins involves statistical comparisons of pairs of price series that should be connected by the marketing system, and it applies to interrelated markets (over space), degrees of product processing (in form), and periods of storage (over time). In using price data to determine the degree of correlation in price movements, it is essential to be clear about the level in the marketing chain to which the prices apply. So-called farmlevel prices, for example, might have been calculated from other prices in the system. The prices must pertain to a comparable quality and form of product, for example, milled rice of a particular variety with a specified percentage of brokens. Prices do need to be reliably collected, and actual prices prevailing in markets are to be distinguished from officially announced prices.

Low correlation coefficients mean the markets are unconnected by actual movements of commodities from one town to another. High correlation of price series between markets, indicating strong price integration, can result from several factors. First, stable prices in all cities can cause high correlation simply because little price movement was observed. Second, high correlation among price movements might indicate perfect competition and efficient price arbitrage. Alternatively, the strong correlations might result from monopoly or effective government policy with little actual market connection occurring among towns. Corroborative evidence is needed to understand the actual formation of prices between markets and hence to explain the reasons for high correlation.


Maps of zones of competitiveness associated with the main marketing chains serve to summarize the description of a marketing system. These maps, or the tabular information that go into their construction, reveal the extent to which marketing costs for a commodity, together with the farm costs of producing it, permit merchants in food producing areas to collect, process, store, and deliver foodstuffs competitively-that is, without government subsidy or protection-to principal consuming centers. This mapping can also be used to show how the national food system is linked to international commodity markets. The connection and relationship between international and domestic grain prices are key elements of domestic food policy and are discussed further in this chapter and in chapter 6. The domestic marketing sector provides the facilities and connecting mechanisms that permit a food price policy to function effectively in an international market context.

Comparing the cost of a domestically produced foodstuff delivered to a port city (the farming cost plus marketing cost) with the locally quoted price for export of the same commodity indicates whether that commodity can be exported without government export subsidy. A similar comparison made with the full import price shows whether the locally produced foodstuff can survive international competition in the port city's wholesale markets and subsequently in interior markets where locally produced foodstuffs have more of a competitive cost advantage. If the commodity is not competitive with imports, a government may choose to limit imports with tariffs or quantitative restrictions. Isocost lines (connecting points of equal costs) can be drawn on the map to indicate, at a given world price for a foodstuff, the limits of production areas able to export competitively and of farming regions able to meet import competition without protection. The likely effects of reductions in production or marketing costs can also be examined by identifying how much additional output would be marketed competitively and where those marketings would originate.

The type of information needed to identify zones of competitiveness is shown in table 4-2. Rice production and marketing in one coastal country, Senegal, is compared with that in three interior countries in West Africa-Mali, Niger, and Upper Volta.
Table 13: Net Social Profitability of Rice Production at Alternative Locations in West Africa (francs per kilogram of rice)
Location of domestic rice production and consumption Social returnsa Social costsb Net social profitability
Consumption at the production center in the interior countries (Mali, Niger, and Upper Volta) 99.2 84.0 15.2
Production at the average distance with consumption at the main center in interior countries 97.0 86.2 10.8
Production at the greatest distance with consumption at the main center in interior countries 97.0 89.9 7.1
Production and consumption in Senegal, distant from the port 84.9 84.0 0.9
Consumption at the production center in Senegal 81.2 84.0 -2.8
Production at the average distance with consumption at the main center in Senegal 79.0 86.2 -7.2
Production at the greatest distance with consumption at the main center in Senegal 79.0 89.9 -10.9

a. Social returns are the opportunity cost of imported rice delivered to the consumption location, according to the following assumptions (in francs per kilogram of rice):
Senegal Interior countries
Price of rice, c.i.f. West African port 74 74
Handling and port charges 5 5
Inland transportation and distribution
costs to main consumption center 0 18
Delivered cost of imported rice at main
consumption center 79 97
Various additional transportation charges are incurred in delivering imported rice to alternative consumption locations.
b. Social costs include farm production costs of 67.7 francs per kilogram, costs of collection and of milling with a small-scale huller of 12.0 francs, and costs of transportation to a main consumption center of 6.5 francs.
Source: Data taken from Charles P. Humphreys and Scott R. Pearson, "Choice of Technique in Sahelian Rice Production," Food Research Institute Studies, vol.18, no.3 (1979-80).

Results are reported in social costs, returns, and profits so that the effects of government policies on actual market prices are removed.

Table 4-2 is constructed to illustrate the decreased social profitability of growing rice the closer the production area is to the port that imports rice. With both production and consumption taking place at the production centers of the interior countries, Mali, Niger, and Upper Volta, social returns are 99.2 francs per kilogram (because of the high cost of bringing imported rice to these areas), while the social cost of production is only 84.0 francs (production plus milling costs and a small transportation charge). The resulting net social profitability is 15.2 francs per kilogram. This profitability declines as domestically produced rice must be transported further while imported rice incurs smaller transportation costs. In the interior countries net social profitability of rice production falls to 7.1 francs per kilogram under the least favorable combination of assumptions for producing area and consuming location.

In the coastal country of Senegal, domestic rice production is barely socially profitable under the most favorable set of locational assumptions. All other combinations show negative social returns, which indicate that Senegalese rice production is not competitive against imported rice without government protection or subsidies for the rice sector. The design of alternative protection and subsidy policies, as well as the analysis for determining their impact or desirability, are treated later in the chapter. The possible mechanisms, however, for increasing the private profitability of growing rice in Senegal are apparent from table 4-2. First, and perhaps easiest, imported rice could have a tariff put on it. A duty of 10 francs per kilogram would ensure the private profitability of rice production except in the most remote rice-growing areas in Senegal. Alternatively, subsidies to farmers, perhaps through subsidized fertilizer or credit, subsidies to rice millers, or subsidies to the transportation system, could accomplish the same goal.

Analytical Techniques for Measuring Marketing Efficiency

The scope for government intervention in marketing is determined by the efficiency and costs of performing the basic marketing functions. If high costs exist, government investments can lower them. If serious inefficiency exists, government policies might improve competitiveness or provide direct competitive standards. Either way, the first task for analysts is to identify empirically the high costs or inefficiencies. If analysts were able to examine the detailed records of marketing agents' costs and returns for transportation, storage, and processing, monopoly profits and inefficiencies might show up directly. Such records are extremely difficult to obtain, however, and those that are available may be of dubious accuracy. The indirect approach of examining price formation at various levels in the marketing chain is frequently more feasible and the data more reliable. Where evidence of inefficiency shows up from such price analysis, more detailed checks of books and records might then be very effective in uncovering the source of high marketing costs.

Marketing efficiency can be analyzed by comparing seasonal price rises with the costs of storage and by correlating market prices in different locations. More extensive analysis focuses on the full marketing margin between farmers and consumers. Models of actual margins between product form and location test the efficiency of processing, the direction of market connection, and the size of margin needed to establish a connection. These techniques usually cannot "prove" that price formation is efficient or inefficient, but each can point to more detailed field surveys which will have a high payoff to further data gathering, analysis, and policy insight.


Seasonal price analysis tests the effectiveness of arbitrage over time. Prices of food crops typically follow a seasonal pattern, falling immediately after the harvest and rising thereafter until the next harvest, as farmers and merchants store some supplies to meet consumer demand throughout the year. In a competitive market, the seasonal price rise should just cover the costs of storage, which consist of interest charges on working capital tied up in the form of stored commodities, provision for commodity losses, the cost of labor and facilities used in storage, and normal profits (including risk-bearing).

By comparing monthly price rises with the monthly costs of storage, analysts can test whether there are excess profits in the storage function. The monthly price rises are derived from an index of wholesale prices, which are usually calculated as the average monthly percentage of a twelve-month moving average. This version of price analysis seeks to identify excess profits in the storage cost portion of the marketing margin. Even if data on storage costs are not available, much insight into the time dimension of the marketing system can be gained by contrasting seasonal price indexes for the main food crops.

An example of seasonal price indexes for four commodities in Ghana is shown in table 4-3. The prices of yams and corn nearly doubled from harvest to preharvest, while those for paddy rose 35 percent and for dried cassava 50 percent. The additional information needed to estimate costs of storage is detailed in table 4-4, which contains storage costs for eight months under different assumed pairs of monthly interest rates and commodity losses in storage. It is known from table 4-3 that corn prices on average rose 95 percent during an eight-month season. This seasonal price rise is consistent with storage losses of 20 percent and a monthly interest rate of 5.7 percent-magnitudes that correspond to a cost of storage for eight months of 99 percent of the purchase value. Alternatively, storage losses of 20 percent, an interest rate of 4.7 percent per month, and other storage costs (such as rent on facilities, labor, and profit) of i percent per month also correspond to the cost of storage for this period. Seasonal price rises for corn thus may reflect actual costs of storage although these costs are quite high. Research in South Asia has tended to show significantly smaller on-farm storage losses and seasonal price rises commensurate with these lower costs of storage.


Correlations of wholesale prices can be calculated between pairs of markets to test market integration.
Table 14: Index of Monthly Wholesale Prices of Yams, Paddy, Corn, and Dried Cassava in Atebubu Market, 1965-74
Crop Jan. Feb. Mar. Apr. May. Jun. Jul. Aug. Sep. Oct. Nov. Dec. Range
Yams - - - - - - - - - - - - -
Mean 104 111 109 126 135 139 90 67 66 67 91 95 73
Standard Deviation 11 20 10 19 17 20 28 15 16 15 16 22
Paddy - - - - - - - - - - - - -
Meana 91 93 96 105 111 110 109 109 110 99 85 82 29
Standard Deviation 10 15 5 12 8 11 16 30 13 6 9 11
Corn - - - - - - - - - - - - -
Meanb 97 99 104 125 141 136 102 74 66 78 92 90 75
Standard Deviation 9 8 7 18 14 26 15 10 6 10 11 6
Dried cassava - - - - - - - - - - - - -
Mean 101 103 83 83 97 102 103 120 126 103 95 83 43
Standard Deviation 42 40 13 13 14 18 15 54 51 26 23 12

Note: Index is average monthly percentage of twelve-month moving average.
a. January 1969 to July 1974
b. January 1968 to march 1974
Source: V. Roy Southworth, William O. Jones, and Scott R. Pearson, "Food Crop Marketing in Atebubu District, Ghana," Food Research Institute Studies, vol. 17, no. 2 (1979), p. 180.
Table 15: Estimated cost of Storage for Eight Months at Various Rates of Interest and Storage Loss
Interest rate per montha Storage loss (percent)
30 20 15 5 none
0 30 20 15 5 0
1 54 35 27 14 8
2 67 46 38 23 17
3 81 58 49 33 27
4 96 71 61 44 37
5 111 85 74 56 48
6 128 99 88 68 59
Note: The cost-of-storage figures in the body of the table are percentages of the purchase value of the amounts left for sale at the end of eight months.
a. Calculations assume interest is compounded annually.
Source: V. Roy Southworth, "Food Crop Marketing in Atebubu District, Ghana," Ph.D. Dissertation, Stanford University, 1981.

Results from one such analysis are reported in table 4-5 and mapped in figure 4-4. Wholesale prices of corn were strongly correlated between pairs of markets in Ghana. Where a coefficient of 1.00 would indicate identical price movements in the two markets, more than half the correlation coefficients were 0.85 or above, and nearly one-fourth were at or above 0.90. A simple correlation coefficient of 0.90 means that 81 percent of the variation in one price series is correlated with variations in the other price series.

Because correlation coefficients are influenced by inflation and by very large seasonal price movements, it is best to choose periods for analysis in which inflation was moderate or to correct for the effects of inflation by correlating price changes rather than actual prices. The price changes are found by subtracting each monthly observation from the one preceding it to obtain first differences.

The map in figure 4-4 shows lines drawn between pairs of cities that have corn price correlation coefficients greater than or equal to 0.90. This mapping includes all but four markets, a result which suggests an integrated marketing system for corn during the period tested. These high correlations indicate a significant degree of spatial arbitrage so long as there are no extreme monopoly conditions or effective government controls. If the analysis shows low correlation coefficients, communication and transportation networks may be inadequate for effective integration of markets.

More significant judgments about the efficiency of spatial arbitrage cannot be made with this technique, for relatively small differences in correlation coefficients can reflect highly profitable market manipulations. The real purpose of intermarket price correlation analysis is to demonstrate that a domestic marketing system really does exist and that it serves to connect the food markets of various cities and towns in the country.
Table 16: Correlation of Wholesale Prices among Pairs of Markets, Ghana, 1965- 1972
- Proportion of total coefficients by commodity
Correlation coefficient Yams Rice Corn Dried cassava
0.95+ 0.02 0.04 0.02 0
0.90-0.94 0 0.03 0.20 0.01
0.85-0.89 0.02 0.03 0.43 0.03
0.80-0.84 0.06 0.08 0.17 0.03
0.75-0.79 0.11 0.13 0.06 0.08
0.70-0.74 0.07 0.17 0.01 0.08
0.65-0.69 0.11 0.13 0 0.17
0.60-0.64 0.13 0.09 0.03 0.09
0.55-0.59 0.07 0.07 0.03 0.13
0.50-0.54 0.11 0.03 0.04 0.10
0-0.49 0.30 0.18 0.02 0.23
<0 0 0.03 0 0.06
Total 1.00 1.00 1.00 1.00
Number of markets 16 16 16 16
Number of pairs 120 120 120 120
First quartile 0.70-0.74 0.75-0.79 0.85-0.89 0.65-0.69
Second quartile 0.60-0.64 0.65-0.69 0.85-0.89 0.55-0.59

Source: V. Roy Southworth, William 0. Jones, and Scott R. Pearson, "Food Crop Marketing in Atebubu District, Ghana," Food Research Institute Studies, vol.17, no. 2 (1979), p. 189.

This simple demonstration can have a powerful effect on policymakers who believe that each market is controlled by a protected monopoly reacting only to local conditions.


If farmers, consumers, and policymakers alike think the marketing margin is too large, an obvious question is "how large is it?" The question is deceptively simple, for the measured size of the margin between the farm-gate price and the retail price can change over time as conditions change in the marketing system. In addition, the margin is calculated as the difference between the retail price and the farm-gate price, but this calculation assumes implicitly that the commodity is actually being marketed through the entire chain from farmer to retail consumer. This may be the right assumption during part of the year-for several months after the harvest, for instance. During other parts of the year, however, no commodities may be flowing from farmers to urban centers. During these periods when no market connection exists between rural and urban markets (or at least the connection is not in the regular direction), comparisons of farm prices and retail prices reveal nothing whatsoever of the size of the marketing margin.

Simple attempts to measure the size of the overall marketing margin by

Figure 4-4. Intermarket Correlation of Wholesale Corn Prices

Note: Corn correlations > 0.90.
Source: V. Roy Southworth, William 0. Jones, and Scott R. Pearson, "Food Crop Marketing in Atebubu District, Ghana," Food Research Institute Studies, vol. 17, no. 2 (1979), p. 184.

calculating the difference between average annual retail and farm-gate prices may significant~ underestimate the true costs of connecting these two markets with actual flows of commodities. The alternative is to specify carefully a simple model of market connection and to use monthly or seasonal data to measure the size of the marketing margin.

One such model is illustrated in figure 4-5, where rice prices are measured on the vertical axis and time during a year along the horizontal axis. Both urban and rural rice prices (in milled rice equivalents) are shown, separated by the full costs of transforming paddy at the farm gate to milled rice at the retail stall. These price observations are all within the same month of observation so that no significant storage costs are tncurred. Figure 4-5 is designed to show what happens to the short-run flow of rice from rural to urban

Figure 4-5. Marketing Margins and Connections between Urban and Rural Markets

Source:C. Peter Timmer, "A Model of Rice Marketing in Indonesia," Food Research Institute Studies, vol. 13, no. 2 (1974), p. 151.

areas and the consequent impact on the observed price margin of any interruption of that flow.

At time to the rice harvest has ended, and both rural and urban rice prices begin their seasonal rise. Urban prices are higher than rural prices by the size of the actual marketing margin M. This margin reflects full competitive costs of connecting the two markets with a physical flow of rice. At time t1, the urban rice price reaches a ceiling dictated either by government policy and supplies from a buffer stock or by the availability of imports at price Pu* Consequently, urban prices cease to rise. Costs of storage in rural areas continue to mount, however, and so rural prices continue to rise from time t1 to time t2, when rural prices actually exceed urban prices by the marketing margin M, and supplies begin to flow back to the rural areas until time t3, when the rural harvest begins in the new season. Rural rice prices fall until they reach their seasonal low at time t01, and the cycle begins again. The urban rice price does not begin to fall until time t4, when the rural rice price has fallen below Pu* by the size of the marketing margin M, and hence the two markets are again connected by physical flows of rice.

How can M be measured? Figure 4-5 shows that only during the periods marked A, from to to t1 and from t4 to t01, is the actual marketing margin M reflected in the difference between the urban and rural rice prices. During those times the analyst can observe the costs the marketing system is incurring to connect the rural and urban markets. Naturally, there is no guarantee that these costs reflect efficient marketing or lack of monopoly influence, but they are the actual costs that must then be compared with estimates of efficient costs. However, the fact that the "measured" margin between t1 and t2, or between t3 and t4, is less than M (and may even be negative) is not evidence that the margins from t0 to t1 and from t4 to t01 are excessive. Only the latter measurements have any meaning at all.

Measuring marketing margins in this manner can quickly provide insight into the dynamics and costs of basic food grain marketing. Published or readily available price data can often be used for the analysis. What the analyst must provide is careful thought about how the marketing system actually works and its seasonal variations. This information comes from many sources and especially from field trips to look at markets. Knowing where the prices are collected and how the markets function are important precursors to estimation of even simple marketing models.

International Commodity Markets

Marketing analysts cannot be content when they understand how their domestic food markets work. Nearly all countries are connected either directly or indirectly to international food markets as well. These connections influence domestic price formation, and so they are important to the immediate issues being treated here. International prices are also integral components of the social profitability analysis outlined in chapter 3 and to the issues of food security that permeate the discussion throughout the book. Food policymakers face a basic issue in deciding whether future food supplies should be produced domestically by the country's own farmers or imported in exchange for other goods and services the country can produce more cheaply and efficiently. The answer can be given only in the context of what level of international prices will prevail when the time comes. These prices are highly unstable, and predicting them is no easy task.


At one level, a food policy analyst can find out about grain prices on international markets with a simple telephone call to one of the major grain export firms, which will be only too happy to quote prices, delivery dates, and conditions of payment. In a very real sense, such a conversation reveals the international price for grain. For the food agency that needs 50,000 tons of wheat for sixty-day delivery, the quoted price is the beginning and end of the story.

Other participants in the country's food system, both public and private, have additional concerns, however. Two in particular are important: how long will the quoted price be relevant, and is the trend of real prices (that is, corrected for inflation) up or down? At any given time experts are more or less evenly divided on these questions. For commodities that are actively traded on ftitures markets (such as wheat, corn, and soybeans), the best information available to market participants is reflected in quotations for contracts with specific delivery dates, up to a year or somewhat more into the future. Even for commodities without active futures markets (such as rice), forward contracts are usually available from major suppliers and give some sense of how market participants view price trends for the near future.

For planning horizons of a year or so, these futures and fonvard markets provide the best information available on likely price trends. Of course, they may not be right. But the evidence of the postwar era shows that no country has been able to outguess these markets in a consistent fashion, and most grain importers have paid higher prices than were available simply by using forward contracts and futures markets to hedge against price risks. Considerable savings seem to be available to countries that develop the skills and bureaucratic rules permitting active use of futures markets to provide the lowest and most stable costs of grain for any given overall market environment.

For the longer run, even futures markets are of little help to planners trying to determine the opportunity costs of investments to raise food production or to lower losses in storage or processing. When investments have payoffs several years in the future, some sense of longer-run price trends in international markets is needed. One way to achieve this is simply to plot over time the real prices of wheat, for example, averaged over a five-or ten-year period to eliminate year-to-year variations and see whether the trend is up or down. To most people's surprise, the trend for the past century has been significantly downward for wheat and corn, except for occasional and short-lived spurts upward, as in 1951 or 1973.

An alternative to such a simple approach, but one that must ultimately be consistent with the trends generated in the markets, is to look at the basic supply and demand factors which generate equilibrium prices in world markets. It must be recognized when taking a supply-demand perspective that world markets for grain do not reflect world supply and demand conditions in total, but rather the economic forces that clear a residual market after internal supply has been balanced against demand inside most countries' borders. Only a few countries permit world grain prices to set directly the signals communicated to domestic consumers and producers; of these few, the United States is by far the most important. When the United States permits free trade of grain across ~t5 borders in response to prices foreign consumers are willing to pay, which is most of the time, then its major grain markets, especially the Chicago grain markets, are the world markets. That is where international grain price formation actually takes place.

The longer-run price trends generated in these markets depend on whether supply curves reflecting available export supplies are shifting outward faster or slower than demand curves. The relevant demand curves reflect desired (and affordable) purchases by countries whose domestic production is insufficient to meet domestic needs within their actual price environment set by trade and subsidy policies. Consequently, the most useful analytical framework for understanding grain price formation on international markets is the same type of excess supply and excess demand curve framework presented in figure 4-2.

In such a framework the protectionist policies of the European Economic Community or the fluctuating grain needs of the U.S.S.R. to meet livestock feed requirements can be incorporated directly into the location and shape of the world market supply and demand curves. Consequently, the analysis can reflect policy environments as well as the longer-run trends in population and income growth on the demand side and technological change, area expansion, and weather and climate on the supply side.

Figure 4-6 illustrates one of two alternative international grain market environments. In this example demand forces are rising more rapidly than factors shifting out supply curves, and the trend of real prices is upward. Such a scenario was widely accepted in the late 1970s in the aftermath of the world food crisis of 1973-74. It would represent a shift in long-run historical trends, however, especially a dramatic slowing in the rate of technical change in agriculture and in area expansion. Both changes are entirely possible, of course, and a slower growth in new areas brought under cultivation is likely. However, if both technical change and area of cultivation (particularly irrigated area) are to some extent a function of earlier incentives to grow more food, then the high prices of the mid-1970s may have been partly self-correcting in the long run through both shifts in the supply curve and in the elasticity of supply itself. As the more elastic, dashed supply curves in figure 4-6 show, even when the shifts are identical to shifts in the inelastic supply curves, the greater elasticity of supply response prevents prices from rising as rapidly.

The second possibility is illustrated in ftgure 4-7, which shows supply curves shifting out more rapidly than demand curves. As was noted, this is consistent with historic patterns but may not reflect future trends if demand pressures increase more quickly because of population and income growth or if supply curves do not shift out as rapidly as has been the case historically. The important role of demand elasticities is illustrated in this figure. If food grain demand is highly inelastic when supply shifts out rapidly, then the trend of real prices is sharply lower. However, if grain demand is more elastic, even

Figure 4-6. International Grain Market Price Formation: Demand Curves Shift Faster than Supply Curves; Prices Rise over Time

Figure 4-7. International Grain Market Price Formation: Supply Curves Shift Faster than Demand Curves; Prices Fall over Time

with exactly the same shifts in demand, then grain prices do not fall so sharply as more demand is induced by the lower prices.

The elasticity of demand for grain is connected to its price level because more end uses become economical as prices drop. Although the distinction between food grains, feed grains, and industrial raw materials is commonly understood, the distinction is basically one of price. The elasticity of demand for wheat, for example, varies with price levels. Grain will be used primarily for direct human consumption at high prices and exhibit a very low demand elasticity. As figure 4-8 shows, at lower prices grain will be fed to animals and eventually even be used industrially. Generally speaking, the lower the price of a commodity or the higher the price of its substitutes, the more likely there can be end-use substitutions. New technical processes, specific government policy, and relative prices of other commodities all influence the degree of substitution. High prices for corn, for example, would shift the feed use portion of the demand curve for wheat, while sharply higher petroleum prices might add to the demand for wheat for energy use.

Figure 4-8. Effects of Changing End Uses on Demand Elasticities

From decade to decade, international price movements are driven fundamentally by world supply and demand forces. This "longer run," however, is composed of a series of annual "short runs," which are in turn affected by other causal variables. These short-run forces tend to be less stable and predictable than population growth or technical change. Moreover, this annual price variability is frequently large enough to obscure the longer-run trends in prices.

Poor weather is one factor that often accounts for substantial leftward shifts in export-supply curves or rightward shifts in import-demand curves. This is especially true when abnormal weather affects a country that is a major actor on either the supply or demand side of an international market. The other significant destabilizing force in world markets is government policy change in one or more key Countries. Unexpected Changes in domestic policy-embargoes, export bans, special barter arrangements, and so forth-often accentuate price effects induced by weather or other short-term variables.

International trade in rice presents a good illustration of all these points. Rice is traded primarily among Asian countries. Bad weather in Asia often affects the rice production of both exporters and importers at the same time. In figure 4-9 poor weather, such as a bad monsoon that affects several countries in South or Southeast Asia, simultaneously causes a leftward shift in the export-supply curve and a rightward shift in the import-demand curve. The consequence is very high prices in years of poor weather and low annual prices in years of good weather in Asia.

Figure 4-9. Effects of Weather on Price Fluctuations in the International Rice Market

Weather-induced variability in rice prices is Compounded by two other sets of factors. First, rice is traded in a "thin" market-a market where only a small percentage of production enters international trade. With world production at approximately 350 million tons (milled rice equivalent) and trade at only 12 million tons, the international market can be extremely sensitive to small changes. A variation of 5 million tons in rice production in Thailand, a major export supplier, would not affect world rice output very much, but it could eliminate 2 million tons of Thai rice exports and reduce world export supplies by one-sixth.

Second, most governments in Asia try to protect producers and consumers from short-run price instability in the rice market by undertaking actions-such as setting up import and export monopolies, long-term purchase arrangements, barter deals, import subsidies, and export taxes-that have the effect of isolating their domestic rice price from changes in the international price. In most countries, movements in international food prices, especially for basic food grains, are not passed on to consumers or producers. As more domestic markets are isolated from international markets by trade and subsidy policies, the international markets themselves become more subject to wide price fluctuations because of the thinness of export supplies and import demands that are price responsive. These fluctuations thus add to pressures on domestic policymakers to isolate themselves from world markets.


The tendency to protect a country's food producers and consumers from unstable world markets is quite understandable, especially in countries that do not use market Systems to generate price signals or to reflect relative scarcity to producers and consumers. Such autarky has its costs, however, including lower overall levels of consumer welfare (although some countries may achieve a more equal distribution of important commodities through such policies). Slow growth in agricultural production is also a common result because signals are missing as to what should and should not be grown. Even in a context of nontrade, the opportunity costs involved are of considerable interest and importance, and these depend on international prices. The important question that remains is what international price should be used as the standard of reference for social profitability analysis or for analysis of various price and trade policies? There will be different answers depending primarily on the time frame of analysis, and an important issue is how to distinguish short-run price fluctuations from long-run trends.

For countries actively trading in international commodity markets, the question has a day-to-day immediacy. Short-run international price fluctuations can be buffered to provide a more stable decisionmaking environment for domestic producers and consumers. But such buffering, whether done through physical storage of grain or trade policy and subsidies, is extremely expensive if domestic prices remain out of line with international prices for Very long. At some point-and the point depends on financial and logistical flexibility~omestic price policy faces strong pressures to conform to the opportunity costs of the outside world. Otherwise budget subsidies play havoc with fiscal policy, and the dynamic distortions become entrenched in the domestic economy.

To follow international market trends as an element of domestic food price policy, three components must be distinguished: the short-run trends over several months that influence import purchases; the medium-run trends that affect consumer and producer prices; and the long-run trends that affect government investment decisions with regard to agricultural infrastructure and research. Analysts can identify these trends by plotting moving averages of real (deflated) prices for the relevant period for decisions. Three-month, two-year, and five- or ten-year moving averages can provide simple and mechanical, yet persistent, reminders that the long run is made up of a sequence of short runs.

These three trends reflect the different time horizons of government decisions that are closely tied to the world price. Whether to use grain from a buffer stock rather than to import grain is a relevant decision over a two- or three-month period, but probably not for a six-month period and almost certainly not for twelve or eighteen months. Similarly, consumer and producer prices can be stabilized for a year or perhaps two by using budget subsidies or import duties to counteract movements in world prices, but the budget costs mount and the distortions begin to get locked into production and consumption patterns before very long. Long-run government investment decisions need a long-run perspective.

A food price policy that actively uses world markets thus requires a sequence of increasingly tighter links as the time horizon of decisions gets shorter. This calls for a complex and somewhat muddy arrangement. Managers of food logistics agencies should often be using an entirely different set of price signals from that used by their own investment departments planning new marketing facilities. Market operatives who are busy buying and selling grain to maintain a buffer stock can be enforcing floor and ceiling prices that are different from either the short-run import price or the long-run investment price. Flexible financing arrangements and a carefully managed buffer stock permit these multiple links between domestic and international prices to serve national food policy objectives.

To further these objectives, two types of analysis are needed. The first looks at the costs and benefits of price policies that use trade barriers or budget subsidies, or both, to place a wedge (or even an iron door) between domestic and international prices. The second is an analysis of policies aimed at narrowing marketing margins, especially dampening seasonal movements in prices and reducing geographical differences. These policies are often implemented with government-operated buffer stocks supplied by imports. Consequently, an understanding of efforts made to narrow marketing margins also provides insights into the potential usefulness of links between domestic and international prices.

Domestic Markets and Price Policy

"Food prices are too high." "Crop prices are too low." Both complaints are heard in virtually all countries. All consumers would like food prices to be lower, to take a smaller portion of their family budgets. All farmers would like their crop prices to be higher, to provide them greater return for their effort and investment. The tension between the two, the food price dilemma, inevitably focuses the attention of consumers, producers, and policymakers alike on the margin between farm prices and consumer prices. All these groups point to the middleman and say "marketing costs are too high."

Several factors common to all food marketing systems lead to this impression, whether it is true or not. First, the marketing system is the narrow point in an hourglass-shaped distribution pattern which first concentrates the crop sales from millions of farmers and then disperses the food to millions of consumers in the time, place, and form in which they want it. Politically, millions of farmers or millions of consumers are forces to be reckoned with; the hundreds or thousands of middlemen usually are not.

Second, operating as a middleman is a very risky business even in developed countries where information is excellent. In developing countries the risks are even larger. Only the highly skilled can survive for very long in such circumstances, and the economic return to these skills is large. In short, many middlemen are quite rich, in sharp contrast to the poverty of both their supplying farmers and consuming households. It is a very short step between the observation of rich middlemen and the conclusion of high-cost, inefficient, and monopolistic marketing. The conclusion, however, is wrong as a logical necessity, and, in a broad array of developing countries, wrong as a matter of fact.

A third reason for the strong perception that marketing margins in developing countries are too high is that they are high. Marketing costs are high when roads and communications are poor, when interest rates and storage losses are high, and when processing facilities are poorly maintained and operated because of difficulties in obtaining working capital or spare parts. In other words, marketing margins are high because the real costs of marketing are high. It is not a matter of short-run private inefficiency and monopoly profits accruing to a few greedy middlemen. Significant opportunities exist for socially profitable investment in the marketing system that will reduce these high costs. The purpose of marketing analysis is to locate these areas of high costs, to identify any inefficiencies and monopoly profits if they exist, and to propose policy initiatives and investments that will lower the real costs of marketing.

Marketing analysis is concerned with both price levels and price margins since the marketing system connects farm prices to consumer prices; in market economies it serves as the arena for price formation at each level. Govern-ments can attempt to set all the important prices to reflect social priorities, or, through a variety of trade and subsidy policies, governments can affect the level of the overall price structure. This analysis is relevant to socialist economies as well as to market economies because allocating resources efficiently and generating and utilizing information are important to both kinds of societies.

The market and welfare aspects of price level interventions are outlined to provide the analyst with a set of tools for assessing the impact of trade and subsidy policies on the welfare of producers and consumers and on the national budget. Determining the welfare impact requires the use of world prices as a standard of reference to judge the opportunity costs of particular policies. Government policies can significantly affect both price levels and margins, and this chapter also presents a wide range of potential interventions that subsidize marketing margins and thereby deal with the food price dilemma.

Price Policy Analysis

In market economies nearly all government economic policies influence the rural-urban terms of trade, defined roughly as the price of food crops relative to the prices of goods and services, including consumer products and farm inputs, purchased by farmers growing food crops. The rural-urban terms of trade is a major factor determining incentives to increase agricultural output while simultaneously signaling consumers about the relative costs of food. For convenience and clarity, the terms of trade will be called the "food parity price." This term is a reminder that farm incentives can be raised in two ways: higher output prices or lower input prices (or lower prices for consumer goods that farmers purchase with their incomes).

The food parity price is principally influenced by two sets of policies: foreign exchange rate, interest rate, and wage rate policies (the macro price policies discussed in chapter 5); and the subsidy and trade policies that drive a wedge between the world price of a product (output or input) and its domestic price. These price policies are illustrated here with food crops in mind, although the reasoning applies equally well to manufactured goods or agricultural inputs, such as fertilizer.

Price policies are judged by their effects on the four food policy objectives-promoting economic efficiency and hence faster growth of income, distributing incomes more equally, guaranteeing adequate nutritional status for all people, and providing security of food supplies. Empirical analysis of a policy requires measurement of the size as well as the likely direction of its impact. In addition, the weights given by governments to the several objectives and the constraints on choice of policy, including international repercussions, determine the actual feasibility and efficacy of a price policy.

Each price policy uses a subsidy or a trade restriction to cause the domestic producer or consumer price, or both, to differ from the world price. A simple consumer subsidy causes both producers and consumers to face lower prices than those in the world market. A specific producer price subsidy can raise the farmers' decision price above world levels while leaving the consumer price at the world price. In the absence of specific policy intervention, the domestic price and world price for a commodity will be the same for both producers and consumers.

A price policy intervention has an impact on the four food policy objectives in the following ways: it affects economic growth by the extent of efficiency losses; income distribution by the direction of income transfers; food security by increases or decreases in quantities traded internationally; and nutritional status by the income transfers price policies effect to or from consumers. (The differential impact on poor consumers of food price changes is not captured within this analytical framework.) The effect of the policy on the food parity price itself depends on whether the policy is applied to a food crop or to a manufactured product purchased by farmers and on whether the policy raises or lowers the domestic price of the commodity. A subsidy on rice imports lowers the food parity price by reducing the food price. A subsidy on fertilizer raises the food parity price by lowering the cost of an important input for food crop production.


A consumer subsidy on importables-goods for which domestic supplies are less than domestic demand in the absence of price policy-is a common price policy intervention in developing countries. Subsidizing rice imports, for example, causes the domestic price of rice for both consumers and producers to be less than the world price. As a result, the quantity of rice produced domestically declines, the amount consumed locally increases, and rice imports are greater than before the subsidy was introduced. The government must use budget resources to lower the price of rice, and this makes consumers better off. In this short-run, static model, when consumers gain, however, the producers of rice lose because their production, sales, and proftts are reduced because of the lower price of rice. In effect, producers are forced to transfer income to consumers. A subsidy on rice imports also causes the food parity price to fall because the domestic rice price is reduced.

All four effects of subsidy policy~n quantities, transfers, efficiency losses, and the food parity price-are shown in figure 4-10. The initial situation, before the subsidy policy, has the domestic price equal to the world price, and so domestic supply is Q1 and domestic demand is Q3, with imports making up the difference (Q3 - Q1). When the government introduces a subsidy on rice prices, thus lowering the domestic price below the world price by PW - Pd demand increases to Q4 and domestic supply drops to Q2, both because of lower prices. The import gap widens to Q4 - Q2.

Figure 4-10. Effects of a Subsidy Policy on a Consumer Import, Rice

The government must pay a budget subsidy on all imported rice since the world price is higher than the domestic price. The total subsidy is (PW - Pd) (Q4 - Q2), or the per unit subsidy times total imports. This amount shows up in figure 4-10 as the rectangle BEHG. The rest of the subsidy is paid implicitly by farmers because of the lower price they receive. Their income transfer to consumers is equal to the unit subsidy times total production, or rectangle ABGF, plus the profits lost by reducing output, the triangle BCG. It is clear that the total economic costs of the subsidy policy are much larger than just the cost of the budget subsidy itself.

Although consumers clearly benefit by this price subsidy on rice, their total gain is less than the sum of the budget subsidy and implicit transfer from farmers. The difference is due to the efficiency losses caused by the price distortions introduced by the wedge between domestic and international rice prices. In this example, there are efficiency losses in both the producing and consuming sectors.

The production efficiency losses are measured by the dotted triangle BCG. Because domestic resources can be used to produce rice more cheaply than the opportunity cost of imports as long as the domestic supply curve is below the world price, the triangle between the world price, the domestic price, and the domestic supply curve is an area of wasted resources. The cost of this waste is paid by the budget, but no commensurate gains accrue to consumers

On the consumer side, the demand curve represents the price consumers are willing to pay for each quantity; so a lower price produces benefits for consumers who were willing to pay a higher price but no longer have to. This consumer surplus is reflected by the area under the demand curve but above the consumer price. In figure 4-10 the increase in consumer surplus is measured by the quadrilateral figure ADHF. The government budget subsidies needed to move the domestic price from Pw to Pd are greater than the gain in consumer surplus that comes from the lower prices. The consumption efficiency loss is shown by the triangle DEH.

One last lesson is apparent from the figure. The country using subsidies to provide consumers with imported rice at prices lower than those in the international market has reduced the degree of food self-sufficiency achieved relative to what would have occurred with free trade. An entirely different result would come from restricting imports and causing domestic rice prices to rise above world levels. If pursued far enough, a policy of pushing up rice prices could result in self-sufficiency for the country illustrated in figure 4-10. In fact, if domestic rice prices were maintained at Ps, the country would exactly reach self-sufficiency, with domestic rice consumption just equaling domestic production. The point is not that pursuing such self-sufficiency is a good or a bad policy, but rather that any policy debate about food self-sufficiency can be conducted only in the context of the domestic price environment relative to world prices. Self-sufficiency at price Ps could be a bitter policy victory if it reduces food consumption and displaces the production of other important agricultural commodities.


Trade policies that place restrictions on the flow of imports or exports of a commodity can be analyzed by using diagrams similar to figure 4-10. A trade restriction can be applied to either the price or the quantity of the commodity to reduce the amount traded internationally and to drive a wedge between the world price and the domestic price. For imports, the trade policy imposes either a per unit tariff (import tax) or a quantitative restriction (import quota) to limit the quantity imported and raise the domestic price above the world price. Likewise, trade policy for exports limits the quantity exported through imposition of either a per unit export tax or an export quota, and the result is to cause the domestic price to be lower than the world price.

If, for example, a trade policy restricts imports of textiles through imposition of a tariff, producers of textiles gain because the domestic price rises above the world price. In response to higher local textile prices, production expands, consumption declines, and the quantity of imports is reduced. Since the domestic price is raised, consumers transfer income to producers and to the government budget because of the duties paid on imports. As with the consumer subsidy on rice, efficiency losses occur in both production and consumption because the policy-adjusted price is higher than the world price, which represents the actual opportunity costs of imports. A tariff on textiles reduces the food parity price because textile prices rise for farmers, thus increasing the price index of manufactured items purchased by farmers. Hence the relative food parity price falls.


Price policies can be classified into six categories:
Table 17: Price Policies
Policies benefiting producers Policies benefiting consumers
Subsidy policies Producer subsidies on importables Consumer subsidies on importables
Producer subsidies on exportables Consumer subsidies on exportables
Trade policies Restrictions on imports Restrictions on exports

Each price policy can be analyzed graphically (as in figure 4-10) to determine the impact on quantities produced, consumed, and traded; on income transfers among producers, consumers, and the budget; and on efficiency losses in production and consumption. The results of these analyses are shown in table 4-6, which summarizes the impact of each price policy and reveals several important effects of price policy interventions.

In the short-run, static world of this analysis all price policy interventions incur efficiency losses in either production or consumption relative to an international price standard. The only exception is when a price policy intervention specifically offsets an existing source of static inefficiency, such as a divergence between public and private profitability that occurs because of externalities or economies of scale. Most policies have efficiency losses in both production and consumption. Long-run efficiency gains are possible if the subsidies call forth dynamic forces leading to technical change or the maturation of infant industries, but these forces are not automatic. In many circumstances the short-run efficiency losses have tended to accumulate and widen the distortions in the economy.

A government uses food price policy to bring about basic changes in the food system. Incentive prices can encourage production, discourage consumption, and reduce imports. Alternatively, food subsidies can reduce production, increase consumption, and increase imports. All price policies have opposite (or at least neutral) effects on production and consumption. To increase both the production and consumption of food

would require maintaining a dual price policy involving subsidies to both producers and consumers.
Table 18: Summary of Efeects of Price Policies
Quantity (increase, no change, or decrease) Transfer (gains, no change, or losses) Efficiency loss (incurred or no change)
Type of policy Production Consumption Trade Producers Consumers Budget Production Consumption
Producer subsidies on importables + 0 - + 0 - x 0
Producer subsidies on exportables + - + + - - x x
Consumer subsidies on importables - + + - + - x x
Consumer subsidies on exportables 0 + - 0 + - 0 x
Restrictions on imports + - - + - + x x
Restrictions on exports - + - - + + x x

Note: X indicates an efficiency loss is incured.

Such combined policies can also be analyzed within the framework used here, but the analysis is somewhat more complicated than the "pure" examples illustrated in figure 4-10 and summarized in table 4-6.

All price policies have an impact on quantities traded internationally since, by definition, the policies apply only to tradable commodities. The welfare effects of price policies for nontradables are difficult to measure because of the absence of an international price standard of comparison. Table 4-6 shows that most price policies reduce trade. This tendency is related to the pervasive efficiency losses incurred by price policies. Because trade leads to gains in economic efficiency through better allocation of productive resources, policies that reduce trade will likely incur efficiency losses.

The allocative effects of price policies on quantities produced, consumed, and traded have corresponding income distribution effects that occur as transfers are made among producers, consumers, and the budget. The full incidence of such transfers can be understood only in the context of the burden of raising tax revenues for the budget, but the direct gains to producers and consumers, before taxes are netted out, are shown in table 4-6. Transfers to producers and consumers tend to mirror the effects on quantities produced and consumed. More important, all subsidy policies incur negative budget transfers while trade restrictions earn the government a budget surplus. Such budget losses or gains are only a part of the total economic transfers occasioned by price policy, and frequently only a tiny part if traded quantities are small relative to total domestic production and consumption. The implicit transfers between producers and consumers are often the most important aspects of food price policy and yet are the least visible.

For some purposes of food policy, knowing the direction of policy effects is enough. In most situations actual measurement is required. Demand and supply elasticities permit the empirical analysis of trade, transfer, and efficiency effects. Because of the static nature of this analysis, these measured effects reflect short-run adjustments only. The dynamic adjustments of the food system to long-run price distortions are at least as important, as is the disaggregated welfare impact relative to the average impact reflected in this analysis. Income and price elasticities for each income class (assembled with the techniques outlined in chapter 2) are necessary to disaggregate the impact on the poor of the various subsidy and trade policies. Typically the nonfarm poor will be disproportionately benefited by price policies that lower food prices and transfer income to consumers at large and will be disproportionately hurt by higher food prices and income transfers to farmers, at least in the short run. Price policies designed for efficiency and more rapid growth of farm output will have a dynamic side effect of creating more jobs in rural areas and probably in urban areas as well.

The policy problem is one of finding mechanisms to protect the food intake of poor urban and rural landless consumers while the dynamic growth process has time to build momentum. The dynamic relationship between an efficient food price policy and performance in the rest of the economy is treated in detail in chapter 5, and the reconciliation of short-run consumer interests with long-run rural productivity is a major element of chapter 6.

While food subsidies that favor consumers can operate within the general price policy environment created by trade and subsidy policies with respect to international border prices, specific subsidies targeted to reach poor urban and rural landless consumers are likely to be implemented through the marketing system. The mutual interaction of food subsidies and the marketing system can be analyzed with the same tools that were used for supply and demand analysis and for determining the effects of trade and subsidy policies.

Subsidies and the Food Price Dilemma

Both socialist and capitalist economies use a variety of subsidies to protect their producers and consumers from the full brunt of the food price dilemma. Developed countries with highly productive farmers often end up paying huge price subsidies to prevent their productivity from driving many farmers into bankruptcy. Developing countries use subsidies to their consumers to allow small household budgets to be stretched just a bit further, thus saving some of the very poor from the brink of starvation itself. Socialist economies often face the food price dilemma directly by attempting to maintain entirely separate and unconnected prices for producers and consumers. This requires that the state carry out all of the functions of the marketing system.

Some countries have managed these respective tasks quite efficiently and have equitable food distribution and vigorous food producing sectors to show for it. Hungary and Costa Rica are examples. Others have been trapped by the size of budget deficits or by the lack of information and the inability to make appropriate allocative decisions. The failure is characteristic of bureaucratic behavior in the absence of markets where price formation takes place. For countries with such problems, the following analysis of marketing subsidies provides insights into the nature of these difficulties. Even for countries that rely extensively on markets to allocate resources and generate price signals, analysis of marketing subsidies can identify cost-effective mechanisms to reach poor consumers.

Subsidizing Marketing Costs

One obvious way to keep farm prices high and consumer prices low is to minimize the marketing margin. Since real economic resources are required to transform food crops in space, time, and form to food that consumers buy and eat, keeping margins below their private costs requires a government subsidy. Most socialist governments and many even in market economies believe that marketing is somehow an unnecessary function. One response is to carry out all marketing tasks directly; another is to legislate narrow (or even zero) margins. Other countries recognize the real value and costs of marketing services and find mechanisms to subsidize the margins as a way of narrowing the price spread between producer and consumer.

The impact of a marketing subsidy depends on whether consumers, producers, or both realize the price benefits made possible by the subsidy. Figure 4-11 illustrates the alternative effects in a simple supply and demand framework. The figure shows a retail supply function for food grain uniformly shifted above the farm supply function by a constant marketing cost equal to Pm - Pf. These are real economic costs incurred in providing essential marketing services between producers and consumers. The retail demand curve for food grain shows the amounts consumers purchase at each price level. The intersection at price level Pm and quantity Qm indicates the market equilibrium in the absence of government intervention or subsidy.

Figure 4-11. Effect of a Marketing Subsidy on Supply and Demand for Food Grains

Source:C. Peter Timmer, "China and the World Food System, " in Ray A. Goldberg, ed., Research in Domestic aud International Agribusiness Management, vol. 2 (Greenwich, Conn.: JAI Press, 1981), p. 111.

A government subsidy on marketing costs equal to Pm - Pf can have a range of results, depending on whether farmers or consumers receive most of the subsidy. Three possibilities are shown in figure 4-11. In the first possibility, farmers receive the entire subsidy so that consumer prices remain at Pm, but farm prices rise to Pm also. Farm output then rises along the farm supply function from Qm to Qm1 and the additional output must be disposed of by the government, usually by subsidizing exports.

Alternatively, consumers receive the entire subsidy so that farm prices remain at Pf, but consumer prices drop to Pf as well. Consumer demand then increases along the retail demand curve to Qf and the government must either ration supplies at price Pf to the quantity Qm that farmers are willing to produce or import additional quantities of food grain equal to Qf - Qm. Whether such imports will require subsidies depends on the relationship between domestic and international prices.

In the third alternative, producers and consumers split the marketing subsidy in such a way that a new equilibrium price and quantity is reached. In figure 4-11 this new position is shown at price Pe and quantity Qe where farmers are just willing to produce along the supply function at price Pe a quantity that consumers are willing to consume at Pe The government must continue to provide the full marketing subsidy of Pm - Pf, but no rationing or subsidized imports or exports are required.

The three examples examined here are taken from a continuum of potential effects of a marketing subsidy. The actual impact will depend on the structure of the marketing sector, how the subsidies are actually implemented, the elasticities of the supply and demand curves, and how carefully the government attempts to regulate the outcome to favor consumers or producers.

The mechanisms by which the government might implement such marketing subsidies are quite diverse. Storage costs can be subsidized with low rental rates in government-owned warehouses or with cheap credit for financing inventories. Gasoline for trucks can be subsidized, and special rates for shipping food crops on a state-owned or regulated rail system can be implemented. Imports of food processing machinery can be subsidized by special tariff and tax concessions and by an overvalued exchange rate or preferential access to foreign exchange. Working capital can be supplied cheaply from the state banking system. All of these explicit or implicit subsidies can be used to reduce the actual costs incurred by the private marketing system.

A more direct subsidy can also be used to lower the marketing margin. Wholesalers, for example, could be paid a per unit subsidy to enable them to sell a particular foodstuff for less than their costs of purchase plus costs of marketing. Rice that cost a merchant Rp 100 per kilogram to buy, store, mill, and transport to the city could be sold for Rp 60 per kilogram if the government provided a subsidy of Rp 40 per kilogram.

Alternatively, the state might simply take over the marketing tasks itself by setting up a parastatal marketing agency with monopoly control over farm sales and consumer purchases. Farm prices and consumer prices thus become a "simple" government policy decision, and a budget subsidy covers any losses on actual operational expenses. When given total monopoly power to handle food commodities, such parastatal marketing agencies have done a rather poor job of defending incentive farm prices and protecting consumer prices without massive subsidies and inefficiencies. When parastatals are used to provide a standard of competitive behavior relative to a private sector, the result is more favorable, although the private sector does tend to have lower real costs of marketing than does the typical parastatal. Consequently, some subsidy is required even in these more limited endeavors.

Both approaches to simple subsidization of marketing margins, either through the private sector or parastatals, run into a very awkward problem. Unless a simple way can be found to separate the "purchase" market from the "selling" market, there is an inevitable tendency for the cheap retail commodity to find its way back to be bought again at the higher farm price. Clearly, this transaction can happen only when the farm price is higher than the retail price for the same commodity, but it is surprising how many governments have tried to implement just such a pricing arrangement. The subsidy burden is quite large when each ton of rice, for example, is subsidized only once. When it appears two or three times on the subsidy rolls, the costs quickly mushroom out of control.

Even success in subsidizing marketing margins does not eliminate marketing costs, for another segment of society is paying them through general tax revenues for the benefit of food producers and consumers. The actual incidence of such a revenue transfer could be socially desirable, but the transfer is still occurring.

The long-run efficacy of subsidized marketing costs is difficult to judge. Future productivity gains are likely to come primarily through more efficient resource allocation led by price mechanisms based on real opportunity costs. If so, state food logistics agencies have a poor record of receiving and transmitting signals of relative scarcity, and better communication of such information will be needed. At the same time, an active government role in dampening the transmittal of sharply fluctuating international prices is likely to be desirable, as is some capacity to smooth out year-to-year fluctuations in domestic price formation.

Subsidizing Poor Consumers

The efficient way to deal with poverty is to transfer general government revenues to poor people and let them make their own allocative decisions about how best to improve their consumption bundle. Few societies wish or are able to accomplish such neutral income transfers, and yet most societies want very much to alleviate the worst manifestations of extreme poverty. Providing direct subsidies to poor consumers in the form of preferential access to or prices for special merit goods-goods whose social value is higher than their market value-is the most common approach in societies as diverse as Sri Lanka, Mexico, and the United States. Because food is considered a merit good in all societies, a number of special food subsidy schemes for poor consumers have been designed and implemented. Chapter 2 categorized these into targeted and nontargeted interventions. Three of the most important targeted food subsidy mechanisms are discussed here: dual price systems, food stamps, and subsidies for foods consumed primarily by the poor.


Several countries with market economies, especially in South Asia, have experimented with dual price systems for basic food grains. Although program details vary considerably, the logic of the approach for a closed economy, that is, one with no food imports, calls for farmers to pay a grain tax based on land cultivated or on historic yields but not on current output. Farmers thus treat the grain tax as a fixed cost of production which does not alter their resource allocation decisions or short-run incentives to produce. The grain obtained from this tax is made available in government-operated or licensed fair-price shops where low-income consumers are permitted to buy a ration quantity at very low prices. Farmers are free to sell their surplus production in an open market where consumers, including the poor, are free to buy whatever quantity they wish at the market-clearing price. Hence, there are two food prices in the system, the cheap ration price set by the government in the fair-price shops and the free-market price set by the equilibrium of supply from farmers and demand from consumers. In some systems the farmers are also paid a low "procurement price" for their grain taken by the state, thus reducing but not eliminating the tax element in the transfer.

The logic and mechanics of this dual price system also apply to a number of socialist economies that maintain parallel rural markets for state purchases and private transactions. Important information is thus generated about relative commodity scarcities, even though state purchases take place at fixed prices. In China these markets are used to provide added sources of income for rural commune members; the price signals indicate to planners how realistic their own purchase prices are for state procurement. Ration shops in cities distribute the procured grain (as well as imported supplies) at low prices to protect the purchasing power of Chinese industrial workers.

The analytics of a dual price system are quite complicated. Food production, consumption by low-income and high-income consumers separately, and an equilibrium supply and demand framework must be integrated into a consistent picture of the food system. The dynamics are illustrated sequentially in figures 4-12 to 4-14.

In figure 4-12, the original supply curve SO and demand curve DO intersect

Figure 4-12. Market Supply and Demand Framework Illustrating the Effect of a Dual Price System

Source:C. Peter Timmer, "China and the World Food System," in Ray A. Goldberg, ed.,Research in Domestic and International Agribusiness Management, vol. 2 (Greenwich, Conn.: JAl Press, 1981), p. 113.

to produce a market price Po and a quantity produced and consumed Q0. The dual price scheme is implemented by government procurement of quantity Qp (in figure 4-13). In the absence of any price response, this procurement of grain supplies shifts the supply and demand curves to S1 and D1, respectively, with the same price and the new market quantity equal to the original quantity less the procurement quantity.

Figure 4-13. Food Production Framework Illustrating the Effect of a Dual Price System

Source: C. Peter Timmer, "China and the World Food System," in Ray A. Goldberg, ed., Research in Domestic and International Agrib~iness Management, vol. 2 (Greenwich, Conn.: JAI Press, 1981), p. 114.

The procurement quantity Qp is sold in fair-price shops to poor consumers at, for example, half the original market price of Po The effect on poor consumers is shown in figure 4-14, where Iop is the original indifference curve for poor consumers before the implementation of a dual price system. After the system is put into effect, poor consumers are able to reach indifference curve Iop by purchasing their entire ration quantities at the ration price, plus small additional quantities of food from the open market at price P2 The low ration price raises the quantities consumed by poor consumers because of an income effect. The added demand shifts overall market demand from

Figure 4-14. High-income and Low-income Consumer Decisionmaking Framework Illustrating the Effect of a Fair-price Shop

Source: C. Peter Timmer, "China and the World Food System," in Ray A. Goldberg, ed., Research in Domestic and International Agribusiness Management, vol. 2 (Greenwich, Conn.: JAI Press, 1981), p. 115.

curve D1 to demand curve D2 (in figure 4-12). The added demand causes supplies to increase along supply curve S1 resulting in a new price equilibrium at P2. A small additional quantity is thus supplied to match the added demand from low-income consumers (and a small decrease in demand from well-off consumers because of the higher price in the market). By assumption, high-income consumers may not buy food at the lower prices of the fair-price shops.

The effects on farmers are illustrated in figure 4-13. The added production is generated by higher food prices, relative to input prices, which cause more intensive use of inputs along the food production function. Total output thus rises from Q0 to Q1 when food prices rise from P0 to P2 (For simplicity, marketing costs are ignored in this discussion.) The government procures a fixed amount of food Qp from farmers, who are free to produce and sell as much as they like in the market. At the original price P0 total production is Q0. At the new market equilibrium price of P2, farmers increase the intensity of cultivation and produce Q1. This increase in output at price P2 must be equal to the increase in demand from consumers through the combined effect of low prices in fair-price shops and higher market price P2 to other consumers.

The effects on low-income and high-income consumers are shown in figure 4-14. The initial quantities of food consumed, Q0 - Qp and Qp respectively, add up to total initial food production Q0. The poor consumer is shown at "starvation" level, while the high-income consumer is well above the "recommended" level of food intake. Both consumers are located on a general income expansion path for food.

After the government procures quantity Qp from farmers, it makes the food available to poor consumers in fair-price shops at a price just half the original price of PO The income effect of this cheap price raises food consumption of poor consumers, with the supplemental quantities being purchased at the new higher market price ~2 The new higher market price P2 forces well-off consumers from indifference curve 10r to 1r1 and their food intake is lowered slightly. The increase in food consumption among the poor minus the small decrease among rich consumers must equal the increased quantity of food produced by farmers as a result of the new higher price P2 The supply and demand framework of figure 4-12 must be consistent with the production function results of figure 4-13 and the consumer decision-making results in figure 4-14. Although these shifts are difficult to demonstrate graphically, markets automatically make the necessary adjustments to reach consistency.

The three figures can also be used to show the impact on poor consumers who might be excluded from the fair-price shops. The new higher market prices may force them below the starvation level. In addition, if fair-price shops do not limit the quantity for sale or restrict their sales to low-income consumers, procurement quantities will be inadequate to stock the shops. Food imports will be needed to fill the gap. Otherwise the fair-price shops may be sold out and not provide low-cost food grain on a regular basis.

The successful dual price system has several key ingredients that are identified by this analytical approach. First, it requires access to significant quantities of low-cost grain from farms large enough to produce sizable marketings. Second, it requires careful control over access to the cheap grain available in the fair-price shops. For the system to work, the ration quantities must be limited to the amounts available, and the rations must be restricted to those at the bottom end of the income distribution, possibly by choosing commodities only the poor will consume. If some of the poor are excluded from the system, they are doubly disadvantaged, for not only are they denied cheap grain from the fair-price shop, but the free-market price is now substantially higher than it would have been in the absence of the dual price system.

Although it might appear that a carefully designed system can operate without a subsidy, especially if the procurement price is sufficiently low (even zero) so that revenue from ration sales pays for procurement and distribution costs, the system clearly requires resource transfers from farmers to consumers. If the burden of the transfers is on large wealthy farmers to benefit low-income consumers, income distribution may become more equal. However, much experience points to the burden falling primarily on low- or middle-income farmers to the benefit of middle-income urban consumers-typically a regressive income transfer. The very poor are frequently excluded from the benefits.

A very carefully designed procurement program can minimize the disincentive effects on agricultural production and hence prevent large efficiency losses, but most existing programs have been neither designed nor implemented so carefully. A common outcome is that prices for all farm output are depressed through the procurement program and incentives are sharply diminished. Even for the most carefully designed program, the tax equivalent of the grain procurement reduces savings available for private, farm-level investment in raising agricultural productivity. Such farm-level investment, when permitted, typically has a high payoff.


Food stamps have been widely used in the United States as the main government program for reducing hunger among poor people. Nearly 20 million people received a net value of over $6,000 million in food stamps in 1981. This large government program remains very controversial, and widely publicized reports of fraud and cheating as well as evidence of signiftcant disincentives for food stamp recipients to take low-paying jobs have eroded public and congressional support. Strongest support for the program comes from labor and social welfare groups and farm lobbies. Farmers have found that the food stamp program adds to demand through the regular food marketing system and thus contributes significantly to farm incomes.

Only a few developing countries have attempted to subsidize their poor consumers by using food stamps-Sri Lanka, Trinidad and Tobago, and Colombia. Despite the theoretical efficiency of food stamps in providing food subsidies targeted precisely to those most in need, the actual implementation record so far is quite mixed. If no serious attempt is made to implement a means test, then food stamps transfer commodity-specific income to a broad range of consumers. If poor consumers are already allocating much of their budgets to these commodities, the transfer serves as a general income transfer rather than as a food-specific transfer. Of course, without a discriminating means test, a large proportion of the population may try to use the program.

Where serious attempts are made to limit food stamps to the most impoverished households, all the problems of implementing an honest and efficient means test arise. Many relatively well-off households slip into the system, many of the most destitute fall outside, and the bureaucratic costs become very large. Food stamp programs as an efficient targeting mechanism for food subsidies can probably be used effectively only in middle-income countries with a skilled civil service and accurate statistical records on at least the urban population. For poorer countries and in the rural areas of even the middle-income countries food stamps are not likely to be effective.


The poor in most societies eat different foods from those consumed by middle- and upper-income groups. As chapter 2 showed, even in those countries where 70 to 90 percent of calories come from starchy staples, the diets of the poor are remarkably different from the average availability shown in a food balance sheet. Poor people's foods tend to be root crops (cassava, sweet potatoes, Irish potatoes) or coarse grains (corn, sorghum, millet, and others). The preferred staple in most societies is either rice or wheat although corn is preferred in some African and Latin American countries. In rice cultures wheat is sometimes regarded as an inferior good.

Such sharp contrasts in food consumption patterns by income class within a country are not caused by differences in taste but by economic necessity. The poor in Indonesia eating cassava and corn would prefer to eat rice, as would the barley eaters in the Republic of Korea in the early 1970s. If a society does not have the bureaucratic and financial resources to provide subsidies for the more expensive preferred foods, subsidies to poor people's foods can be effectively self-targeting. If only the poor choose to eat the subsidized inferior staples, only the poor capture the subsidy.

At the same time, many of the inferior foods are produced by very poor farmers on marginal lands at considerable distance from major urban centers. Marketing subsidies that raise the returns to these farmers while lowering the costs to the consumers may work simultaneously on both dimensions of poverty. Simply forcing down prices, however, would have a devastating impact on the incomes and welfare of some of the poorest of the rural poor.

Such subsidies have both short-run and long-run costs. In the short run, implementing subsidies for commodities that do not travel or store well (root crops) or for which well-developed marketing systems do not exist is usually not feasible without significant investment in food technology and improved marketing infrastructure. Because farmers tend to switch to growing more profttable crops, obtaining supplies of these commodities when market prices are being forced down is an obvious problem. Providing subsidies to poor consumers via commodity-specific food stamps (available with minimal bureaucratic processing) while allowing incentive prices for farmers in commercial markets might be more feasible. Alternatively, an imported commodity can be an efficient carrier of subsidies to poor consumers and have less impact on domestic farmers. A subsidized low-quality wheat flour might have this effect in Sri Lanka, for example.

The longer-run effects are more troubling and suggest that disaggregated commodity price policies probably can serve only as short-run bridges across the food price dilemma. The distortions introduced by significant subsidies on a single commodity can eventually be very powerful. Low prices for high quality wheat at one time led to almost one-third of Sri Lanka's calories being provided by a foodstuff it could not grow. Livestock industries find heavily subsidized corn or wheat a cheap, high-quality animal feed. Such grain-fed livestock industries redirect the subsidy from poor people to rich people. The lower prices for these inferior foods almost inevitably dampen incentives for research and development of new technology for the crops and reduce the profitability of growing them.

The implementation of food subsidies for poor consumers through the regular channels of the marketing system is the most efficient way to protect food intake of the poor when price incentives to farmers are improved. But both the analytics of the design and the historical implementation record suggest great difficulties in isolating the targeted food subsidy programs for poor people from more general food subsidies for all consumers, or all urban consumers. Such general subsidies have enormous fiscal effects and serious disincentive consequences for agriculture. Both the problems and the potential rewards of successfully implementing targeted food subsidy programs explain the extensive attention in this book to these topics.

Government Intervention and Policy Perspective

Developing a marketing strategy for government intervention requires a clear vision of what the marketing system should accomplish in the future plus a suitably detailed empirical understanding of what it actually does accomplish now. This chapter has attempted to provide both the vision and the analytical tools to address the empirical questions of market performance. The elements of a marketing strategy can be identified by combining these two approaches.


Marketing analysis shows that the formation of price margins is largely a function of two elements: the costs of transportation, storage, and processing; and the efficiency with which these marketing services are provided. The government's role is to invest in the components of a marketing system to the extent that social benefits from lowered marketing costs match the social opportunity costs of the public resources needed for the investment. Lowering marketing costs is clearly a good thing as long as more resources are saved than are needed to save them.

In addition to a concern for lowering the real costs of marketing, governments need to focus on the efficiency with which marketing services are provided. In market economies, inefficiency means excess profits, and excess proftts mean monopolistic middlemen or collusion in price formation. Both sources of excess proftts are extremely difficult to regulate directly because of enforcement problems. In the face of solid evidence of market inefficiency (as opposed to high costs), governments are faced with two quite divergent alternatives. The first is to improve the competitiveness of the marketing system by creating better market access for potential participants who might provide marketing services and by distributing better information for consumers, producers, and marketing agents about factors likely to affect price formation.

The second alternative is for the government to provide the marketing services directly, setting a competitive standard that all other marketing participants must meet. In most countries that have followed this course, it was the government which could not meet the competitive standard of the existing marketing participants. Continuing the government's marketing role thus would require either significant budget subsidies to cover the high costs or the banning of private market activities to eliminate the competition, thereby forcing consumers to pay the costs of government inefficiency.

Banning private marketing activities never improves the welfare of broad groups of either farmers or consumers. No government has ever been completely successful with such a ban, although many have tried. Subsidized government marketing agencies, however, can play an important and socially profitable role. If the subsidy is not too large, the agency can reflect a competitive standard for private marketing agents without driving them out of business. At the same time, a public marketing agency can implement a price stabilization policy that requires active government intervention to defend a floor price for farmers and a ceiling price for consumers-a partlcularly useful role for a government agency in reducing extreme and unexpected seasonal price swings. Again, the margin between the two will determine the amount of budget subsidy required by the public agency to cover its own real costs as well as the ability of the private trade to continue to provide a signiftcant share of marketing services. Forcing the entire food marketing burden into government hands either by active policy or by default presents an awesome task of coordination, physical handling, and price formation. No government has handled the task on its own even when it wanted to, and those that have tried to handle more rather than less have achieved less rather than more.

The purpose of having efficient and low-cost marketing services provided to the food system in particular, and to the whole economy in general, is twofold. First, and most important in the short run, low marketing costs are the most efficient and sustainable solution to the food price dilemma. The narrower the margin because of genuinely low marketing costs and highly efficient price formation, the more consumers and producers both can share in the productivity potential of a healthy agricultural economy.

The second purpose is to enable markets to function in their dynamlc role of coordinating resource allocation and providing accurate signals to producers and consumers that reflect the opportunity costs of their decisions. Implicit in this role is the ability of producers and consumers, and indeed of the marketing agents as well, to react quickly and efficiently to new price signals from the market. Although most economic models show decisioninakers moving from one equilibrium to another equilibrium after, for example, a price Change, the actual process of adjusting to a new disequilibrium environment is much more complicated.

Economic development is inherently a process of continuing disequilibrium. The economic value of being able to cope successfully and efficiently with disequilibrium rises sharply as new technology, new markets, and new opportunities are thrust into traditional economies. T. W. Schultz has emphasized how important education is in building the capacity to process new information and decide quickly on appropriate responses. Education, at least at the level of functional literacy, is considered by most governments to be a basic human need. Beyond this, however, education is essential to the dynamic efficiency of market systems and their distributional outcome. Poor people tend to be the least educated and thus have the least capacity to respond appropriately to the opportunities offered by disequilibrium situations.


In a competitive food marketing system the level of prices is determined simultaneously with the various margins among prices. Because real economic resources must be used to provide marketing services, the formation of price margins is the major determinant of the efficiency of resource allocation in the marketing sector. Similarly, the formation of price levels determines the efficiency of resource allocation in food production and consumption. These price levels are also crucial determinants of income distribution, especially between urban and rural sectors, and of the distribution of food intake.

Many governments in both socialist and market systems have intervened in the formation of price levels to influence the distribution of income and food and have willingly sacriftced efficiency goals to do so. Through trade and subsidy instruments that most governments can implement quite effectively, food prices can be set (at least for short periods) with wide discretion.

The tension between the desire to set food prices for short-run distributional reasons and the need to avoid long-run productivity losses that mount from such seriously distorted prices is not easily resolved, even with collective ownership and decisionmaking. The dilemma is not so sharp for wealthy societies which have the budgetary and management resources to use food stamps or other welfare programs to protect the consumption levels of very poor people. For less fortunate societies more structural compromises have seemed inevitable, with at least part of the price and market system being used to deliver food to the poor.

To reconcile the conflict between food prices set for efficiency purposes and food prices set for distributional purposes, ways must be found to target the effects of price policy interventions. Food stamps target benefits to poor people very effectively if the bureaucratic capacity exists to identify and reach them with appropriate amounts of stamps. This is a big "if." Fair-price shops in the context of a dual food price system have been tried extensively in South Asia. Research by the International Food Policy Research Institute points to positive consumption consequences of the systems in both Bangladesh and India. The programs, however, have had very large leakages to nonpoor consumers and significant disincentive effects for farmers. Most rural consumers have been out of reach of the system. Price targeting by commodity, with subsidies paid only for food consumed primarily by the poor, has not been tried extensively anywhere, but the strategy must face problems of supply and the potential use of subsidized inferior foods as livestock feed. All three approaches used simultaneously in an intersecting targeted program-food stamps to gain access to fair-price shops that sell foods consumed primarily by the poor-might offer a financially feasible and effective alternative.

The temptation for governments to intervene in food marketing and price formation is very great. A variety of interventions can contribute to important social goals. Investments can lower marketing costs. Well-managed public buffer stocks can improve price stability and set competitive standards for private markets. Appropriate price levels can improve production incentives or increase food consumption. But an even greater variety of interventions can disrupt the food system or bring it to a standstill. Narrow margins set by legislation can drive most food marketing activities into hiding. Heavy consumer subsidies for the major foodstuff can distort producer incentives a~d place enormous burdens on the budget. Parastatal marketing agencies with monopoly power can immobilize the efficient allocation of resources and simultaneously damage the welfare of both poor farmers and poor consumers.

Marketing systems are at once fragile and robust. They are fragile because government actions can, with ease, drastically raise the risks of storing, transporting, and processing food. With higher risks come higher marketing costs and a distortion of price signals to producers, consumers, and marketing agents alike. Poor societies can ill afford the waste from such distortion and inefficiency, especially when caused by the actions of their own governments.

The robustness of marketing systems comes from the flexibility, adaptability, and sheer drive to survive and do better that is characteristic of most decisionmakers in most marketing systems. Markets never entirely disappear. They are too important for both people and society. A strategy that nurtures their development by encouraging equal access, rather than forces them into hiding, is likely to pay handsome social rewards.

Bibliographical Note

Discussions of markets and prices and of price policies can be found in a number of books on microeconomic theory, but no textbook deals with these topics explicitly in the context of food crop marketing in developing countries. The two texts that best Complement this Chapter are Raymond G. Bressler, Jr., and Richard A. King, Markets, Prices, and Interregional Trade (New York: John Wiley, 1970), and William G. Tomek and Kenneth L. Robinson, Agricultural Product Prices (Ithaca, N.Y.: Cornell University Press, 1981). Analysis of price formation in the presence of divergences between private and social valuations in supply or demand is presented in W. M. Corden, Trade Policy and Economic Welfare (Oxford: Clarendon Press, 1974). A modern statement of trade theory and a complete discussion of international price formation are contained in Richard E. Caves and Ronald W. Jones, World Trade and Payments: An Introduction (Boston, Mass.: Little, Brown, 1981). An important but quite sophisticated analysis of international buffer stock schemes to stabilize commodity prices is presented in David M. G. Newbery and Joseph E. Stiglitz, The Theory of Commodity Price Stabilization: A Study in the Economics of Risk (Oxford: Clarendon Press, 1981). Their conclusions cast serious doubts on the workability of and benefits from such schemes. This reinforces this chapter's emphasis on domestic price policy and the use of international markets to further domestic objectives.

A discussion of market power and price formation is in Kenneth Arrow, "Toward a Theory of Price Adjustment," The Allocation of Economic Resources (Stanford, Calif.: Stanford University Press, 1959). This volume also contains a classic article by Hendrik Houthakker on seasonal price formation, "The Scope and Limits of Futures Trading." Further discussion of market failure is in Francis Bator, "The Anatomy of Market Failure," Quarterly Journal of Economics (August 1958). The role of education in coping with disequilibrium is discussed in Theodore W. Schultz, lnvestment in Human Capital: The Role of Education and of Research (New York: Free Press, 1971).

Examples of the analytical methods introduced in the chapter are presented in a variety of books and articles. Techniques of price analysis are explained in Frederick V. Waugh, Demand and Price Analysis, Technical Bulletin no.1316 (Washington, D.C.: U.S. Department of Agriculture, Economic and Statistical Analysis Division, 1964). An empirical analysis of seasonal price formation is presented in Richard H. Goldman, "Seasonal Rice Prices in Indonesia, 1953-69: An Anticipatory Price Analysis," Food Research Institute Studies, vol. 13, no. 2 (1974), pp. 99-143. A study of marketing margins is contained in C. Peter Timmer, "A Model of Rice Marketing Margins in Indonesia," Food Research Institute Studies, vol. 13, no. 2 (1974), pp. 145-67. A discussion of range of farmer choice in crop sales is in Ammar Siamwalla, "Farmers and Middlemen: Aspects of Agricultural Marketing in Thailand" (Bangkok: United Nations Asian Development Institute, 1975).

Techniques that can be used to analyze proposed projects for expanding marketing capacity are discussed in J. Price Gittinger, Economic Analysis of Agricultural Projects, 2d ed. (Baltimore, Md.: Johns Hopkins University Press, 1982). The analysis needed to generate data for mapping marketing margins is illustrated in Scott R. Pearson, J. Dirck Stryker, Charles P. Humphreys,and others, Rice in West Africa: Policy and Economics (Stanford, Calif.: Stanford University Press, 1981). An analysis of the benefits of price stabilization is presented in Saleh Afiff and C. Peter Timmer, "Rice Policy in Indonesia," Food Research Institute Studies, vol. 10, no. 2 (1971), pp. 131-59. The use of rice prices as a major policy instrument by Asian governments is analyzed in two special issues of Food Research Institute Studies edited by C. Peter Timmer, "The Political Economy of Rice in Asia," vol.14, nos. 3 and 4 (1975).

Various approaches have been used to describe and analyze marketing systems and policies for food crops. Elements of the perspective used here draw on the work of William 0. Jones, especially Marketing StapLe Food Crops in Tropical Africa (Ithaca, N.Y.: Cornell University Press, 1972). A case study using the Jones approach is reported in V. Roy Southworth, William 0. Jones, and Scott R. Pearson, "Food Crop Marketing in Atebubu District, Ghana," Food Research Institute Studies, voL 17, no. 2 (1979), pp.157-95.

The issues of postharvest losses during storage, transporting, and processing are treated in a major section of Nevin S. Scrimshaw and Mitchell B. Wallerstein, eds., Nutrition Policy Implementation: Issues and Experience (New York:Plenum Press, 1972), which includes a comment by Michael Lipton that cites important work done at the Institute of Development Studies at Sussex. See, for example, Martin Greeley, "Appropriate Technology: Recent Indian Experience with Farm-level Food-grain Research," Food Policy, vol.3, no. i (February 1978), pp. 39-49.

Three review papers place many of the food crop marketing issues discussed in this chapter in development perspective: John C. Abbott, "The Development of Marketing Institutions," in Herman M. Southworth and Bruce F. Johnston, eds., Agricultural Development and Economic Growth (Ithaca, N.Y.:Cornell University Press, 1969), pp.364-93; Vernon Ruttan, "Agricultural Product and Factor Markets in Southeast Asia," Economic Development and Cultural Change, vol.17, no.4 (July1969), pp.501-19; and Barbara Harriss, "There Is a Method in My Madness: Or Is It Vice Versa? Measuring Agricultural Market Performance," Food Research Institute Studies, vol. 17, no. 2 (1979), pp.197-218.

Book-length case studies include: Paul J. Bohannon and George Dalton, eds., Markets in Africa (Evanston, Ill.: Northwestern University Press, 1962); Uma J. Lele, Food Grain Marketing in India: Private Performance and Public Policy (Ithaca, N.Y.: Cornell University Press, 1971); and Leon A. Mears, Rice Marketing in the Republik of Indonesia (Jakarta: P. T. Pembangunan, 1961). Results from a number of West African studies are contained in Elliot Berg, "Marketing, Food Policy and Storage of Food Grains in the Sahel" (Ann Arbor: University of MichiganlUsAID, Center for Research on Economic Development, 1977). A series of Latin American studies is summarized in Kelly Harrison and others, "Improving Food Marketing Systems in Developing Countries: Experience from Latin America," Research Report no.6(East Lansing: Michigan State University, Latin American Studies Center, November 1976).

Trade protection has been an important topic of recent research. D. Gale Johnson's book, World Agriculture in Disarray (New York: Macmillan, 1973), documented many of the inefficiencies in agricultural trade patterns and their high cost to consumers in rich and poor countries. Alex McCalla and Timothy Josling, eds., Imperfect Markets in Agricultural Trade (Montclair, N.J.: Allenheld-Osman, 1981), has an extensive bibliography and assessment of research needs. Studies by International Food Policy Research Institute, Agricultural Protection in OECD Countries: Its Cost to Less Developed Countries (Washington, D.C.: IFPRI, December 1980), and Jimmye Hiliman, Nontariff Agricultural Trade Barriers (Lincoln: University of Nebraska Press, 1978), look specifically ac the problems of developing countries caused by the agricultural trade barriers of developed countries.

[Top] [Prev] [Next] [Bottom]

Copyright © 1983, The World Bank All rights reserved.