IX. Price Subsidies
- Ke-young Chu, and Richard Hemming
- Published Date:
- September 1991
Under what circumstances are subsidies justified? How is their impact on resource allocation, distribution, the budget, and the balance of payments assessed?
What alternative forms can food subsidies take? Which form is most effective in helping the poor while containing budgetary cost?
Is there a case for a general subsidy on energy products in developing countries, or should it be restricted to products such as kerosene, which are used more by the poor?
What are the arguments for fertilizer subsidies? Why is there resistance to reducing such subsidies?
A commodity subsidy is created when, as a result of public policy, the price paid by the consumer of a good is lowered below or the price received by a producer is increased above what it would otherwise be in the absence of the policy. Subsidies may be either explicit, taking the form of payments by the government to producers or consumers showing up in the budget, or they may be implicit, and entail no apparent budgetary cost. In centrally planned and many developing economies, in addition to budgetary subsidies, subsidies take the form of sizable positive gaps between the world market price and the domestic price of a product. In developing countries, the most common type of explicit budgetary subsidy is arguably that paid by the government on basic foods such as rice and wheat or flour, or energy products such as kerosene. Often, the subsidy will be paid to a marketing agency owned or controlled by the government that is required to sell its commodities at a price less than that at which the goods were procured. In industrial countries, producer subsidies for various agricultural commodities are common.
Subsidies can take many forms. Most of the goods and services directly provided by the government are subsidized either by providing them free of charge, as with primary and secondary education, or at prices that are substantially below costs. Tax expenditures are another form of subsidy common in industrial countries. In particular, owner-occupied housing is often indirectly subsidized through the income tax system by deducting mortgage interest payments and excluding capital gains on the sale of a taxpayer’s principal residence from taxable income.
Arguments for Subsidies
Broadly speaking, there are two basic rationales for commodity subsidies: first, they are a means of redistributing income, and secondly, they can be used to offset market failures of one kind or another. As regards their redistributive aspect, commodity subsidies are second-best instruments of redistribution in countries where the tax and transfer system is not a practical means of redistribution. Subsidies can be viewed as negative taxes, and rates (some of which may be negative) should be set to equalize the marginal social cost of revenue across commodities. If society values extra income for the poor more highly than for the affluent, then commodities that account for a relatively large share of the budget of the poor, or commodities on which the poor spend more than the rich, are appropriate candidates for a low rate of tax or subsidization. The ideal good for a redistributive subsidy should satisfy both these conditions, but finding such a commodity may be difficult.
Market failure describes situations where the level of production or consumption of a good or service would be inappropriate without the government’s intervention. For example, it has been argued that subsidies on fertilizers that reduce their price below the free market level are justified because farmers in many countries lack the information necessary to make a fully informed appraisal of the potential impact of fertilizer on crop yields. The absence of well-functioning rural credit markets is cited to justify the provision of subsidized agricultural credit. Public transportation is implicitly subsidized when user charges are set at levels below the cost of provision, and this practice is justified by the argument that it reduces the costs of road congestion entailed by excessive use of private transportation. As a final example, a subsidy on basic foods has been justified on the grounds that it enhances the productivity of the work force through its beneficial impact on nutrition.
The Impact of Subsidies
An understanding of the allocative impact of a subsidy scheme is critical to any appraisal of its impact on the economy. The analysis of the allocative impact of a subsidy becomes a complicated matter, however, when the subsidy is large enough that its effects spill over into other markets. Consider the case of a subsidy on a consumer good, where a marketing agency processes that good and sells it for less than the cost of production and distribution. The subsidy lowers the price of the good, which will normally increase the quantity demanded. If the domestic producer price is controlled and not allowed to increase, the additional quantity demanded must be supplied by imports, because there will be no incentive to increase domestic production. Reliance on imports will be even greater if the government tries to finance the subsidy to consumers by lowering the producer price.
This strategy has been adopted for agricultural commodities by governments who have assumed that the domestic supply response to producer price changes is negligible. While the short-run price elasticity of supply of some goods may not be large, often it is likely to increase over time as resources are reallocated from the production of other goods, so that governments resorting to this strategy find their reliance on imports growing or exports declining.
The allocative effect also depends on the impact of the subsidy on demand for the subsidized commodity and other commodities. A general subsidy—one where there is no limit to the quantity of purchases that is subsidized—will have both an income and a substitution effect. By lowering the relative price of the subsidized good, it encourages the substitution of that good for other goods. The income effect will also boost the consumption of the subsidized good and other goods, unless the good is inferior.
By contrast, when a subsidy applies to only a limited quantity of purchases—as in the case of a rationed commodity—no substitution effect exists when a household is already consuming more of the good than the amount to which it is entitled under the rationing scheme. These inframarginal subsidies are equivalent to an income transfer equal to the amount of the subsidy per unit times the quantity of rationed purchases. In these cases, the subsidy will not lead to a substantial increase in demand for the good unless the subsidy represents a substantial share of the average consumer’s income and the subsidized good has a high income elasticity of demand.
The allocative effects of a subsidy will also depend on the way it is financed. When a government has many different tax instruments at its disposal, the introduction of a subsidy can in principle lead to changes in tax rates. The allocative effect of all these changes would have to be taken into account, and it would be hazardous to consider in isolation the impact of the subsidy. If, however, the subsidy is financed by an increase in the rate of a broad-based tax such as a general sales tax, the allocative effects on markets other than the market for subsidized good are likely to be diffused, and can probably be ignored.
There is an obvious difficulty with the use of consumer subsidies as an instrument of income redistribution, in that it is not easy to limit them to goods that are consumed primarily or exclusively by the poor. This does not mean that generalized subsidies cannot be an effective means of redistributing income, but it does imply that to have an effect on the poor a subsidy program must be relatively large, implying significant transfers to the nonpoor. Marketed goods with a negative income elasticity (i.e., inferior goods) are ideal candidates for a redistributive subsidy. One such good found in a number of West African countries is cassava. In other countries, however, it may not be possible to identify an inferior good that makes up a significant part of the budgets of poor households. Alternative means of targeting subsidies at the poor are explored in the note on Poverty and Social Security.
Fiscal and trade effects
There is no reason in principle why a subsidy program should entail a budgetary problem, but generalized subsidies for staple products have sometimes caused difficulties. One reason for this is that subsidies are created by the operations of state marketing agencies that sell the subsidized commodities at prices that are not adjusted regularly for increases in the cost of production. In this case, when there is no mechanism in place to put a cap on the subsidy per unit of product, the subsidy is said to be open-ended. Such subsidies have explosive potential, and there are many instances where they have risen sharply and then become entrenched. Failure to adjust the regulated price can also reduce domestic supply, and lead to an increase in imports. These imports will often be more expensive than local production, and thus lead to an increase in the subsidy. Under these conditions, food subsidy programs will be vulnerable to a bad harvest if domestic stocks are low. The failure to adjust prices increases the temptation to move the subsidy off budget or to finance it through an implicit tax on producers in the form of producer price controls. The cost of imported food can be disguised by adopting a special exchange rate for food imports. The subsidy then shows up as a loss of the central bank (see the note on Fiscal Activities of Public Institutions).
A subsidy that takes the form of a fixed payment per unit of the subsidized good is not affected by rising costs. However, this type of subsidy is not common. If producers are instead subsidized through direct payments, the effect on the market price is uncertain and depends on the elasticities of supply and demand for the subsidized commodity. If supply is relatively inelastic, most of the subsidy is a transfer to the producer. Direct payments to consumers lower the price effectively paid for the product, but these are both difficult and potentially costly to administer, because the consumer’s purchase of the subsidized commodity must be validated. These real or apparent drawbacks with direct payments may explain why state marketing boards are heavily involved in the administration of commodity subsidies. Governments often want to exercise direct control over the prices of certain staples, and this control generally is not achievable through direct payments.
Food subsidies of one kind or another are common in developing countries, and in a number of cases have accounted for a substantial share of government expenditure. Perhaps the best known case is that of Egypt, where food subsidies represented between 7 to 15 percent of total public expenditure between 1973 and 1981. The subsidy for the rice ration scheme in Sri Lanka, which up to 1977 was available to the whole population, accounted for 11 percent of total expenditure in that year. Other examples include the general subsidy for wheat in Morocco—a producer subsidy which was in effect during much of the 1980s—and the subsidy for maize in Zambia.
Food subsidies may be either general or restricted in some way. In addition to the often substantial direct cost of the subsidy, a general subsidy may entail additional expenditures by the public sector for storage and distribution. These expenditures arise because the government, to ensure that the price is less than the unregulated market price, will have to intervene in the procurement and distribution network. Its intervention is usually carried out through a state-owned marketing agency, which is responsible for procuring the subsidized crop or food from local producers and reselling it to wholesalers or retailers. Where support prices for agricultural products are maintained above world prices, a producer subsidy results, which may be compounded by a budgetary subsidy to ensure prices lower than world prices for consumers.
Public ownership is not essential, however, if it is possible to subsidize the private distribution network. Payment of the subsidy near the stage of first sale is usually the most feasible way of implementing it, because this stage of the distributive chain is likely to comprise relatively few establishments. It will, however, be necessary to ensure that the subsidy is ultimately passed on to the final consumer. The increase in the quantity of food demanded requires that local producers be paid a price higher than would prevail in an unregulated market, unless the increase in the quantity demanded is to be met through an increase in the volume of imports. If the increase in demand is to be met at least in part by an increase in domestic production, a general subsidy, if explicit, would normally entail a subsidy to both producers and consumers, and the price received by the producer could exceed the price paid by the consumers. The resulting two-tier market would create substantial opportunities for fraud, which may explain why countries come to rely on increased imports to satisfy the increase in demand entailed by the subsidized final price.
Food subsidies and malnutrition
Over and above their impact on income distribution, food subsidies can be used to increase the intake of nutrients among the malnourished, and to lessen the incidence of malnutrition, and thereby enhance labor productivity. Viewed from this perspective, generalized subsidies have been supported as a sound food policy tool on the ground that they increase the probability that the malnourished receive adequate food. In part, this maximization of coverage may result from the absence of stigma attached to purchases of foods that are subject to a generalized subsidy. In addition, with generalized subsidies both income and substitution effects act to increase the quantity of food demanded. Because both income and price elasticities of demand for staple food tend to be high among the poor, their food consumption will increase. However, there may be substitution of subsidized for nonsubsidized food, so that the increase in consumption of the subsidized food will overstate the increase in caloric intake.
The clear disadvantage of a generalized subsidy is its lack of precision—many or most of those who benefit from it are not malnourished, and its broad coverage constrains the average subsidy per unit of food. Nonetheless, a general subsidy can increase the overall progressivity of a country’s tax and transfer system, depending on the changes in tax and other subsidy rates associated with its implementation. When viewed in isolation from the impact of such changes, several subsidy programs in developing countries have been found to entail an increase in the real income of the poor by 15–25 percent.
From the perspective of nutrition policy, a general subsidy may not be the most cost-effective policy because a low intake of food may not be the principal cause of malnutrition. The Egyptian food subsidy contributes to a relatively high level of consumption of food among the poor, but the incidence of malnutrition is significant because of diseases caused by inadequate sanitation. There is little direct evidence as to whether subsidies improve nutritional status significantly, although there is some evidence that they lead to an increase in children’s weight. An additional drawback to such subsidies is that, despite their name, general subsidies can have an urban bias because distribution outlets will tend to be located in larger towns and cities. If the subsidy covers only transport and distribution costs, however, the final price to the consumer will differ little from the rural price. Providing grains at below-market prices in the producing areas is undesirable because it distorts incentives and creates opportunities for profiteering.
Subsidies on Energy Products
Subsidies on energy products are often found in developing countries, especially in oil exporting countries. These subsidies are often implicit—for example, they may take the form of below-world-market prices for the products of the national petroleum company, in which case they need not figure in budgetary subsidies, and their effect is to lower profits of the petroleum company below what they otherwise would be.
Cross-subsidization is also a common phenomenon, and in many countries the petroleum companies will administer a fund that compensates producers of subsidized products from special levies on other products. Often kerosene and heating oil are subsidized by levies on regular or high-octane gasoline. Rarely are all subsidies in the energy sector included in the budget. Subsidization of energy products is by no means the universal rule, however; in countries that are net importers of energy products, these products may bear very high rates of tax. Subsidies on energy products can result from efforts to insulate domestic prices from temporary fluctuations in world market prices. Such price stabilization programs will result at a minimum in periodic subsidies, which can become permanent if domestic price adjustments tend to lag increases in world market prices.
Arguments for energy subsidies
Subsidies for energy products are usually justified on either distributive grounds or because they promote industrial development. The sharp increases in petroleum prices in 1973-74 and again in 1979-80 led to an increased degree of subsidization in many countries for these reasons and from a more general concern that adjustment to a drastic change in the price of a major commodity should be gradual. In many oil exporting countries, the existence of subsidies is probably explained by the political impossibility of not subsidizing a product that the country appears to have in great abundance. The notion that an exhaustible resource may be really more expensive than the costs of extraction and distribution alone is one that carries little weight with the public at large.
As noted earlier, the use of subsidies as a redistributive tool is more easily justified in developing than industrial countries. The argument for general subsidization of energy products on distributive grounds in developing countries, however, is not as strong as it might appear. The poor are not necessarily heavy direct or indirect consumers of such products as high-octane gasoline. The subsidization of public transport may be a more efficient way of increasing the real incomes of the poor than a subsidy on high-octane gasoline, because only the very affluent tend to own automobiles. Certain products, such as kerosene, may be more suitable for a redistributive subsidy if used for heating and cooking in relatively poor households.
The argument for subsidization of energy products to promote industrial development is really quite weak. Even if industrial promotion is considered desirable, the subsidization of energy products is a highly indirect and inefficient means to this end: it subsidizes energy-intensive industries more than other industries and promotes the use of energy-intensive production techniques. This is inappropriate in countries which are net importers of energy products. There are additional and more urgent reasons for avoiding the subsidization of energy products. Their use generally entails costs to society in the form of environmental degradation and congestion of public transport facilities, and extra taxes on these products may be the most efficient way of internalizing these costs (see also the notes on Public Expenditure and the Environment and Pricing and Cost Recovery).
Fertilizer subsidies are also found in many developing countries. They are relatively easy to administer when fertilizer is either imported or produced by a small number of domestic enterprises. In many countries, fertilizer subsidies have been used to encourage use of high-yield varieties of rice and wheat.
There are several possible rationales for such a subsidy. First, purchases of fertilizer are typically concentrated in a relatively short period of the crop cycle, and capital market imperfections may mean that farmers are unable to finance the necessary outlay. As an instrument to compensate for capital market imperfections, this subsidy is clearly second-best. Second, subsidies may be justified if farmers are unaware of the technical properties of fertilizer or more recent varieties of fertilizer: the demonstration effect means that the costs of promoting fertilizer use exceed its value to the farmers who first benefit, but fertilizer use by other farmers is encouraged. The argument based on imperfect knowledge and the demonstration effect implies that over time the subsidy on given varieties of fertilizer should be reduced and eventually eliminated. Any assessment of the benefits of a subsidy should bear in mind the risks of overfertilization, when fertilizer is applied to the point where its cost at the margin exceeds the value of the additional output it produces.
Fertilizer subsidies have also been used to offset the impact of price controls on farm output on the supply of agricultural products and the incomes of farmers—China is a prominent example—and as a redistributive device. As a means of reducing the disincentive effects of price controls on agricultural supply, the subsidy is inefficient, because it affects only one element of costs and encourages overfertilization. They are also ineffective as a redistributive tool, because they tend to subsidize farmers with larger holdings of land, but have little impact on the subsistence farmer. Hence, there is considerable resistance to their removal from better-off beneficiaries. Moreover, it may not be possible to eliminate or even significantly reduce a subsidy on fertilizer as long as agricultural product prices are controlled.
The Indian fertilizer subsidy
Subsidies are an important item in the budget of the central government of India; in recent years their share of total expenditure and net lending has approached 10 percent. The largest single commodity subsidy is that for fertilizer. In 1988/89, it accounted for 4 percent of total expenditure and net lending and represented 0.8 percent of GDP. The justification for the fertilizer subsidy is that insufficient use is made of “non-traditional” factors of production, and the subsidy encourages the use of a productive input whose marginal cost remains below its marginal return.
Expenditure on the fertilizer subsidy has grown rapidly during the 1980s, mainly because of a failure to adjust the farmgate price of fertilizer to reflect increases in the cost of production. Until 1974, the Government set a single country-wide factory price and then established the farmgate price by adding margins to cover the costs of freight and distribution. As a result, changes in the cost of manufacturing were reflected in changes in the farmgate price. The enormous increase in the price of feedstock in 1973/74 led to a change in this policy, and the link between factory and farmgate prices was broken. One farmgate price is still set for the entire country—it differs somewhat from the price actually paid by farmers, which includes an additional margin to reflect local costs—but this is no higher in nominal terms than it was in 1981, and has therefore fallen substantially relative to costs.
In 1977, a substantial change was also made to the way in which factory prices are determined. This price—called the retention price—is now set for each factory individually, on the basis of an estimate of average unit costs at a specified capacity utilization rate, which currently ranges from 80 to 90 percent. This estimate is premised on the assumption that the given capacity utilization rate for each factory is achieved at minimum cost in terms of the employment of labor and use of feedstock and other productive inputs. The retention price is set so that the return on net worth is 12 percent.
Two separate issues raised by the Indian fertilizer subsidy need to be considered: (i) Is the amount of the subsidy justifiable as a correction for market failure or some other distortion?; and (ii) Is the subsidy mechanism efficient? With regard to the first issue, it is noteworthy that the increase in the per unit subsidy for fertilizer that has taken place since 1980 has been associated with an increase in fertilizer consumption of over 50 percent. It may be that the subsidy results in additional use of fertilizer on land where its marginal productivity is still high enough to cover its real cost, but with the passage of time this becomes increasingly less likely. This would suggest that the subsidy should be declining over time. As regards the second issue, there can be little dispute that the mechanism of the subsidy is inefficient. If each fertilizer plant expands its production to the point at which its marginal cost equals its (unique) reference price, then marginal costs will differ from one factory to the next, and a basic condition of efficiency in production is violated. This would not be the case if one reference price were set for all producers. The reference price mechanism also destroys the incentive for new entrants to the market to adopt the best-practice technology.
Gittinger, J. Price, Joanne Leslie, and Caroline Hoisington, Food Policy-Integrating Supply, Distribution, and Consumption (Baltimore: Johns Hopkins University, 1987).
Pinstrup-Andersen, Per, ed., Food Subsidies in Developing Countries—Costs, Benefits, and Policy Options, (Baltimore: Johns Hopkins University, 1988).
Schneider, Robert R., “Food Subsidies—a Multiple Price Model,” IMF Staff Papers, 32 (Washington: International Monetary Fund, 1985).