Although a tariff system confers economic protection, it can be a mixed blessing for domestic producers. In the first place, exporters generally suffer from a tariff regime, since it raises the costs of their inputs, directly and indirectly, but it does not allow their export price to rise correspondingly. This was well documented for the case of Colombia in a classic study by Jorge García García.14 Secondly, even an import-competing sub-sector can be hurt by a tariff regime, if tariff rates are not uniform and they are higher on its inputs than on the products that compete with its outputs. In analytic terms, this effect is captured by measuring effective protection rates as opposed to simple (nominal) protection rates, and effective rates can be negative if protection is higher on inputs than on outputs. Thirdly, it is well known that tariffs at significant levels can undermine the competitiveness of domestic industries and sectors, as the additional economic profits delivered by the tariff protection tend to weaken the resolve to increase productivity.
For this last reason, it now is an accepted principle that tariff rates should not be high in general, and if they are, a program should be put in place to scale them downward over time. Usually, free trade agreements incorporate clauses to put these objectives into effect. As noted, in the case of NAFTA, up to 15 years have been allowed to eliminate some agricultural tariffs, but an agreement for their eventual elimination has been sealed by a treaty.
In addition to avoiding high tariff levels, the second basic principle for tariff systems is that their rates be relatively uniform over sectors and products. For the sake of fostering economic efficiency, this principle is exceedingly important. Three exceptions to it are discussed below, but otherwise the more uniform the tariffs are, then the better for the sake of promoting economic growth. Uneven tariff protection favors some industries or sub-sectors over others, and often those which are so favored turn out to be the least competitive over the longer run. They may be the industries which feel the pinch of competition the most and therefore have applied the greatest political pressure for protection. This is an example of the rent-seeking behavior induced by subsidies (implicit subsidies in this case) that was mentioned in Chapter 3. Empirical documentation of the strongly inverse relationship between the competitiveness of products and their rate of tariff protection has been developed for Honduras where, for example, coffee, one of the most competitive products, has received negative economic protection and sugar, one of the least competitive, has received highly positive protection.15
In essence, a system of tariffs that are unequal across products places the government in the position of 'picking winners', and experience has amply demonstrated that governments are much less able to do that successfully than the market is. A common variant of a non-uniform tariff system is a set of graduated tariffs, with the lowest rate reserved for primary products, a higher rate applied to industrial intermediate goods, and the highest rate applied to manufactured final consumption goods. Such a system was applied, for example, in the Republic of Korea in the early decades of its economic take-off, and more recently in Guatemala. This system emerged from the early school of development thought which emphasized industrialization as the road to economic improvement (Chapter 1). It discriminates
14. J. García García, The Effects of Exchange Rates and Commercial Policy on Agricultural Incentives in Colombia, Research Report No. 24, International Food Policy Research Institute, Washington, DC, USA, 1981.
15. See Roger D. Norton and Magdalena García U., Tasas de Protección Efectiva de los Principales Productos Agrícolas, Serie Estudios de Economía Agrícola No. 4, Proyecto APAH, Abt Associates Inc., Tegucigalpa, DC, Honduras, May 1992, p. 22 and tables.
against agriculture. If a country has a comparative advantage in some agricultural products, it may be asked, why should the tariff system be designed to penalize those products and implicitly subsidize industry?
Provided that tariffs are in the low-to-moderate range, it can be said that it is at least as important, from an economic development perspective, to make them as uniform as possible, as to continue reducing them. In order for a country to exploit its comparative advantage to the fullest, and hence maximize its growth prospects, it is very important to align domestic relative prices with international relative prices as much as possible. A uniform tariff policy marks a major step in this direction.
Another common practice is to deviate from uniformity by setting tariffs on basic food products at zero, either by granting tariff exemptions to a State importing agency or by simply legislating the rates at zero. The purpose of such a policy is to make foods such as dairy products and cereals more affordable for the poor, but it can be argued that tariffs are not the most appropriate instrument for achieving this aim. In the first place, subsidizing foods through tariff exemptions means that farmers bear the entire burden of the subsidy, and thus domestic production is likely to decrease relative to imported sources of supply. In the words of Valdes, 'A predictable and well-documented result of cheap food policies is that self-sufficiency in the commodities in question, that is, those that are implicitly subsidized, decreases rapidly'.16
Secondly, this kind of subsidy is completely nontargeted, so that all families receive it in proportion to their levels of food consumption, regardless of their incomes. In that sense, it usually turns out to be a regressive subsidy. The
The beneficial effect for the rural poor, including the landless, of rising food prices was found in Honduras by Dean Schreiner and Magdalena García. The devaluations carried out in stages between 1988 and 1990 had a marked positive effect on real agricultural prices. The production response was immediate and sustained. All rural income strata benefitted, and the poorest strata benefitted the most. (See the study by D. Schreiner and M. García, Principales Resultados de los Programas de Ajuste Estructural en Honduras, Serie Estudios de Economía Agrícola No. 5, Proyecto APAH, Tegucigalpa, DC, Honduras, June 1993.)
value of the subsidy is greatest for the most well-off families. For example, in Kenya it was found that 'subsidies on sifted maize meal in urban Kenya were disproportionately captured by the more well-off strata of the urban population'.17
Thirdly, tariff exemptions on food usually exacerbate the problem of rural poverty. This effect occurs because reducing production incentives affects producers of all farm size classes and accordingly also reduces employment for rural landless laborers. In most lower-income countries, the bulk of the poverty is found in rural areas, so the problem is a serious one from a national perspective. Exceptions to this conclusion may occur when the proportion of rural landless to rural landed is very high, as in Bangladesh.18
A recommended alternative to eliminating tariffs on food imports is a program of targeted
16. Alberto Valdes, 'Explicit versus Implicit Food Subsidies: Distribution of Costs', in Per Pinstrup-Andersen (Ed.), Food Subsidies in Developing Countries: Costs, Benefits and Policy Options, The Johns Hopkins University Press, Baltimore, MD, USA, 1988, Chapter 5.
17. Reprinted from Food Policy, 22(5), T. S. Jayne and G. Argwings-Kodhek, 'Consumer response to maize market liberalization in urban Kenya', p. 456, Copyright (1997), with permission from Elsevier.
18. See Raisuddin Ahmed, Foodgrain Supply, Distribution and Consumption Policies within a Dual Pricing Mechanism: A Case Study in Bangladesh, Research Report No. 8, International Food Policy Research Institute, Washington, DC, USA, 1979.
food assistance, funded through fiscal channels.19 In this way, the beneficiaries are only, or almost only, the poor, while those who bear the burden of financing the programs are taxpayers in general. If the taxation system is even modestly progressive, both the benefits and cost side of such programs will have a progressive incidence on the well-being of the population.
The political calculus for eliminating tariff exemptions on foods can be complicated by the interests of agro-industries which rely on imported inputs, such as the poultry and feed concentrates industries, which often use imported yellow corn and sorghum, and industries which use powdered milk as a raw material. Therefore, it is important to point out clearly to decision-makers that such tariff exemptions constitute non-targeted and regressive subsidies, perhaps more so if they represent responses to pressures from agro-industries. (Some countries have negotiated tariff-quota agreements between agro-industry and farmers associations, under which tariffs fall drastically, sometimes to zero, after all the domestic crop is sold.)
A complicating factor is that the agreements for international food assistance programs such as PL 480 generally prohibit the imposition of border taxes on the products they supply. However, in Honduras it was found to be possible to impose a tax on the PL 480 products after the government imported them and when they were sold to domestic agro-industries, thus compensating for the tariff exemption.20
In short, there are several valid reasons for making tariffs as uniform as possible and for not providing tariff exemptions on food. In addition to the considerations cited above, moving toward a policy of more uniform tariffs usually is beneficial for agriculture, since the biases in the nonuniform systems generally favor industry. In the case of Brazil, for example, for the period 1966-1983 it was found that tariffs and subsidies for imports and exports of non-agricultural goods had the result that crops were discriminated against in economic terms.21 In the case of Honduras in the mid-1980s, Julio Berlinsky found that the average effective protection rate for industry was 99%, while the above-referenced study of Norton and Garcia found that effective protection for agriculture was nil and even negative for many crops.22
The three justifiable exceptions to a uniform tariff policy are as follows:
(1) The case of international subsidies that reduce world market prices. These subsidies are determined by decision-makers in a few, better-off countries, and the international prices for those commodities reflect those decisions. Given the essentially irreversible nature of the flow of labor
19. For a guide to such programs, see Margaret Grosh, Administering Targeted Social Programs in Latin America: From Platitudes to Practice, The World Bank, Regional and Sectoral Studies, Washington, DC, USA, 1994.
20. According to Section 401(b) of the PL 480 legislation, called the 'Bellmon Amendment', there must be assurance that the delivery of commodities under the program does not create disincentives for production in the receiving country. Nevertheless, it is logically impossible to satisfy both this requirement and the provisions of Sections 103(c) and 103(n) which refer to the 'uniform marketing requirements'. The latter require that PL 480 shipments be in addition to what the country normally would have imported; however, if this is the case, then clearly such shipments will drive the domestic price below what it would have been otherwise. This point is brought out in R. D. Norton and C. A. Benito, 'An Evaluation of the PL 480 Title I Programs in Honduras', a report prepared for the USAID Mission to Honduras, August, 1987. In El Salvador in the early 1990s, the Government terminated the PL 480 program rather than accept the provision of tariff exemptions.
21. Antonio Salazar P. Brandâo and José L. Carvalho, 'Brazil', in A. O. Krueger, M. Schiff and A. Valdés (Eds), The Political Economy of Agricultural Pricing Policy: Volume I, Latin America, The Johns Hopkins University Press, Baltimore, MD, USA, 1991, Chapter 3, p. 67.
22. Julio Berlinsky, Honduras: Estructura de Protección de la Industria Manufacturera, UNDP, Buenos Aires, Argentina, July 1986.
from agriculture to urban occupations (Chapter 2), it is difficult to argue that a developing country should accept relative prices determined in other, more developed countries when those prices may drive its labor permanently out of agriculture into the army of unemployed and underemployed. If the more developed countries eventually reduce their subsidies substantially (which seems likely), then a less developed country would face a daunting challenge in trying to create conditions for a recovery of its agriculture. It would be a very difficult task once a significant part of the rural labor force has left for the cities. In addition to these considerations, international subsidies can worsen rural poverty in low-income countries, and the social and economic costs of excessively rapid rural-urban migration are high, as noted previously, and so it may be asked, why should a poor country acquiesce in policy decisions made elsewhere, when the consequences would include those costs? This is the essence of Peter Timmer's argument, mentioned earlier in Chapter 3.
Sometimes, the question of compensating domestic producers for the 'international subsidies' is viewed through the lens of anti-dumping provisions, in both the WTO agreement and domestic legislation. That legal route is an option. Pursuing it has the aim of gaining internationally recognized legal authorization to apply compensatory tariffs, or 'countervailing duties' as they are called in that context. However, to prove that dumping has occurred is time-consuming and costly, and the process itself gives rise to frictions among trading partners. The procedure is basically designed for cases of firms selling at below cost, rather than to subsidies provided through national policies. A simpler expedient for a developing nation would be to simply legislate a tariff surcharge (preferably on the base of an otherwise uniform tariff rate) which is equivalent to the international price distortion caused by subsidies to producers in other countries, as mentioned in Chapter 3. The WTO accords stipulate a tariff ceiling ('binding'), in most cases high enough to permit all or most of the international price distortion to be compensated by a surcharge.
If such a policy were implemented, it would be important to base the magnitude of the surcharges on estimates of price distortions that have been made or sponsored by international agencies, and also to legislate a provision which mandates their review at intervals of say, five years. The reviews could gather new evidence from the most recent calculations of the international price effects of subsidies. The surcharge should disappear when the price effects of international subsidies become negligible. In any event, such a policy would be applicable to only a few products, chiefly or entirely agricultural, which were affected by those subsidies.
This option would not benefit a developing country if it did not have a comparative advantage in the products affected by international subsidies, in the absence of those subsidies. In those cases, the developing country would distort its own resource allocation and affect negatively its agroindustries by the higher tariff. (2) The case of price bands for smoothing price fluctuations. Properly designed, price bands are neutral on average with respect to economic protection. However, recently a panel of the WTO ruled that they are in violation of WTO accords in Chile, where they were first applied. The full reason for the ruling is not clear, although the panel may have wanted to ensure consistency with an earlier ruling outlawing the European system of variable levies - even though price bands are different. Given the valuable role that price bands have played in some developing countries, it is worth a brief review of them in the event that their potential use is revived in future trade negotiations, and because they are still in operation in places where they have not been challenged formally. First of all, unlike variable levies, price bands have no link to a domestic support price. They compensate for excessively high peaks in international commodity prices (for the benefit of consumers) and, in equal measure, for excessively low troughs (for the benefit of producers). They accomplish this by varying the tariff rate on, say, a bi-weekly basis, according to an automatic formula which, in turn, is based on the historical series of the relevant international price. When the international price rises above its historical trend line by more than a pre-established percentage or amount (often set as one standard deviation), then the corresponding tariff begins to rachet downward. When the price fluctuates downward by more than the same amount, then the tariff is raised. All movements in the tariffs caused by the price bands are temporary, subject to modification in the following month or two-week period.23
Under the price band system, there is not a fixed floor price or ceiling price, but rather there are threshold prices which trigger tariff changes, and these threshold prices change each month as the most recent month is added to the moving time series of historical prices and an earlier one is dropped from the series.
In the case of the Salvadorean and Honduran policy reforms, implementation of a price band system proved to be a vital element in convincing producers to accept a free trade regime. They had been fearful of the economic damage that could be done by pronounced drops in international prices, even if they were temporary. The weakness of price bands in some cases has been a lack of fully transparent administration. If they were to be revived under international accords, provision should be made for a small panel of international experts to supervise the installation of the system and to monitor its functioning from time to time. A good explanation of how price bands are designed and a summary of Central American experience in operating them, as well as an explanation of some of common misconceptions about the bands, has been provided in two publications by Julio Paz Cafferata.24
(3) When a crop is the main source of food and income for the rural poor. This case is especially relevant to the conditions of recent years, when many agricultural prices have declined continuously and substantially. The reasons why the rural poor may not be able to readily adapt to new occupations or improve their agricultural yields, in the face of declining real agricultural prices, have been mentioned. If tariff surcharges have not been applied to a basic product grown by the poor for reason (1) above, and if a price band system is not functioning, it can be important to set tariffs on the products at or near the levels of the WTO bindings to help alleviate poverty in rural areas. A less distortive measure would be to provide the rural poor with direct income support, but given that they are usually numerous and geographically dispersed, and titles to their farms often are not registered, it is more difficult to target assistance on them than it is for the urban poor.
Only one or two products would merit higher tariffs for this reason, and the higher level of tariffs should have a time limit, as in Mexico. One of the costs of such a policy is that not all of the protection afforded by the tariff would go to the poor. It would be constructive to augment the policy by depositing at least some of the proceeds from the tariff in a special fund for improvement of technology on small farms, and to use those funds for farmer training, yield-enhancing investments, and investments in alternative crops or livestock products. It would also be important to target such assistance primarily on the women in poor families, to the extent feasible. The tariff could be visualized as an adjustment tariff if the revenues from it were effectively used for helping poor farm families make economic adaptations.
23. Price bands should not be confused with the variable levy system initiated by the European Union. A variable levy is designed to continuously ensure that border prices are equivalent to domestic support prices. Under a price band, there need not be a support price, as the movements in the tariffs are not linked to any domestic price, but rather to the historical pattern of international prices. Usually, a 60-month moving average of international prices is used to establish the trend line and to update it continuously.
24. (a) Julio Paz Cafferata, 'El Sistema de Bandas de Precio: Una Alternativa de Política de Precios para los Granos en Honduras', RUTA II, San José, Costa Rica, June, 1990. (b) Julio Paz Cafferata, 'La Experiencia de Banda de Precios para la Regulación de las Importaciones de Granos en Centroamérica', in Mercados y Granos Básicos en Nicaragua: Hacia una Nueva Visión sobre Producción y Comercialización, H. Clemens, D. Greene and M. Spoor (Eds), Escuela de Economía Agrícola y Programa Agrícola CONAGRO/IDB/ UNDP, Managua, Nicaragua, 1994, Chapter 5.
Each of these exceptions to a uniform tariff policy has a clear rationale and they should not be used to justify a policy of protectionism. At most, they would apply to three or four agricultural products, usually one or two. One of the mechanisms, the price bands, does not constitute protection at all over the longer run.
In addition to uniformity over products, subject to these three classes of exception, a third basic principle is that tariff systems should be relatively stable over time, except for downward adjustments in phases that have been programmed years in advance. In practice, this is one of the most difficult principles to gain adherence to, for political leaders sometimes succumb to the temptation to tinker with tariffs in response to perceived crises in industries or special interests. Varying tariffs at frequent intervals is very damaging to economic growth prospects, because it creates a large degree of uncertainty about future economic policy and therefore it discourages productive investments.
In light of these principles, it can be observed that the WTO rules have important weaknesses from a viewpoint of promoting economic development. First, they allow a fairly high ceiling on tariffs during a long transitional period,25 and consequently they allow considerable variation among tariffs by product, with some items located at or near the lower end of the permissible range (which usually is zero) and others at the upper end. While this room for maneuver may be convenient from a viewpoint of imposing compensatory surcharges for a few agricultural products affected by international subsidies, in general it is quite damaging to allocative efficiency and therefore to economic growth prospects. Secondly, the rules do not require stability over time in national tariff systems.
Thirdly, they discriminate against exports, in favor of import-competing sectors, by allowing tariffs to be considerably higher than export subsidies, which, in fact, are supposed to be eliminated. For the poorest countries, a policy which was neutral between the two classes of goods would permit uniform export subsidies in exactly the same percentage as a uniform tariff on imports. All countries will benefit in the long run from a worldwide reduction of tariffs and export subsidies, but during the lengthy period in which they are being phased out, allowing tariffs to be significantly higher than export subsidies introduces a bias that works against economic development for many countries. In light of these observations, although the Uruguay Round and other WTO negotiations have represented important advances toward liberalizing trade, with concomitant benefits for all countries, it can be seen that the WTO rules do not fully take into account the perspective of development policy.
Domestic policy can counteract these weaknesses to a considerable extent, by making tariffs uniform except for the three cases noted above, by forging a national agreement to maintain stability in the system over time, and in some cases by implementing a program of export subsidies up to the 10% limit if tariffs are at that level or higher. Since the aggregate value of imports almost always exceeds that of exports in developing economies (because developing countries are net capital importers), such subsidies could be financed out of tariff revenues, with room to spare.
Was this article helpful?