A quarterly magazine of the IMF

September 2002, Volume 39, Number 3

Poor countries could boost growth and reduce poverty by expanding exports to the rich countries and to each other. But, despite the progress made in trade liberalization under successive multilateral agreements, many barriers persist in both developing and industrial countries.

Living standards in Korea, only 50 years ago a poor country dependent on foreign aid for half its national budget, have been catching up to those in the industrial countries. One of the reasons is a strong export sector that has fueled Korea's economic growth while evolving to keep up with changes in international demand.

Most developing countries, however, unlike Korea, have been unable to overcome the obstacles to expanding and diversifying their exports. The primary commodities on which many rely for export earnings have faced stagnant demand and been battered by volatile prices, and the two sectors in which developing countries have a strong comparative advantage—agriculture and labor-intensive manufactures, like textiles and clothing—are heavily protected not only in the industrial countries but in developing countries as well.

Most quantitative restrictions and other nontariff barriers have been converted into tariffs since the Uruguay Round of trade talks, improving the transparency of trade regimes. Protectionism has actually increased in some cases, however, and trade barriers are still higher for the products typically exported by developing countries than for those from industrial countries. This is partly because developing countries made little effort to participate in multilateral trade talks before the Uruguay Round and partly because of the political sensitivity of liberalizing agriculture and labor-intensive manufactures. Developing countries themselves have high tariffs that limit trade among them. The average tariff in developing countries is 14 percent, and in the least developed countries, 17.9 percent, compared with 5.2 percent in the industrial countries. This article focuses on protection in merchandise trade. The liberalization of trade in services, which is generally subject to far greater restrictions, offers opportunities for developing countries that, according to some estimates, are even greater than in merchandise trade (for instance, in labor-intensive services that require the temporary movement of workers (World Bank, 2002)).

Patterns of protection

Developing and industrial countries both pay dearly for protectionism. Estimates from a variety of sources (in particular, World Bank, 2002) of annual static welfare gains from eliminating barriers to merchandise trade range from $250 billion to $620 billion, of which one-third to one-half would accrue to developing countries. The response of investment and technology to a freer international trade regime would generate additional dynamic gains.

And yet protection persists, in many guises and to a greater extent than is revealed by the customary references to average most-favored-nation (MFN) tariffs. These do not reflect specific tariffs and tariff-rate quotas, trade remedies such as antidumping duties, and the effects of rules of origin and environmental and technical standards. Nor do the averages capture the impact of tariff peaks and escalation, preference schemes, or measures that contribute to the uncertainty of market access and therefore discourage export expansion.

Table 1 presents the combined ad valorem tariff equivalents (AVEs) (import tariffs as a percentage of the value or price of imported products) of various protectionist measures from the perspective of groups of exporting countries. It shows that, while Canadian and European Union (EU) barriers hit low- and middle-income exporters hardest, Japanese (in agriculture) and U.S. protection is highest on the products exported by the least developed countries (LDCs).

Specific tariffs and tariff-rate quotas. These account for a significant share of the AVEs shown in Table 1. A specific tariff, an absolute amount of money charged per unit of imports regardless of the price (for example, €350 a ton on sugar imports into the EU), is generally regarded as less transparent and more distortionary than an ad valorem tariff. With tariff-rate quotas, the tariff is different above or below a specific quantity or value of imported items. Tariff-rate quotas, established under the Uruguay Round, were originally intended to ensure minimum market access for sensitive products. However, out-of-quota tariffs can be prohibitive, and even in-quota tariffs are often high.

Tariff peaks and tariff escalation. Although average industrial tariffs have dropped, between 6 and 14 percent of Quad (Canada, the EU, Japan, and the United States, in World Trade Organization (WTO) parlance) tariff lines are subject to "tariff peaks" (tariffs at or over 15 percent). In Canada and the United States, tariff peaks are concentrated in textiles and clothing, in the EU and Japan, in agriculture, food products, and footwear. Tariff peaks are even more common in developing countries. Estimates suggest that if all tariffs were capped at 15 percent, AVEs on textiles and clothing would drop 20 percent for imports from most countries into the United States and 59 percent for imports from China, while AVEs on agricultural and food products imported by the EU would drop 40-60 percent.

Tariff escalation, which is seen in both industrial and developing countries, is designed to protect a processing or manufacturing industry in the importing country, which sets low tariffs on imported materials used by its industry and higher tariffs on imported finished products that would compete with the domestic industry's own products. This creates hurdles for countries trying to move up the technology ladder, discouraging them from expanding their processing industries and diversifying exports, and therefore leaving them dependent on commodities, whose prices are often volatile.

Contingent protection. This includes trade remedies permitted by the WTO to counter unfair trade practices or manage sudden surges in imports. Most economists agree that antidumping measures, the most widely used, have been abused for protectionist purposes. Over 1,800 antidumping investigations have been initiated since 1995. While industrial countries have traditionally been the main users of such measures, developing countries have been more active in recent years; between 1994 and 2001, they initiated almost two-thirds of all investigations. Developing countries have also been the target of nearly 60 percent of investigations—most of which have been initiated by other developing countries. Most antidumping actions have been concentrated in a small number of sectors, especially steel, chemicals, textiles, and consumer electronics, often at the low-technology end of the product range. Antidumping investigations—or even the mere possibility of them—create a great deal of uncertainty for potential investors in export sectors and discourage exporters from passing on their efficiency gains, thus forcing consumers and downstream industries in importing countries to pay higher prices.

Standards. Standards and regulations play an important role in trade by ensuring the quality, safety, and technical compatibility of products and production processes, but they may at times be more stringent than is appropriate and abused by those seeking to raise the costs of potential competitors. Annual notifications of new technical barriers to the WTO and its predecessor, the General Agreement on Tariffs and Trade (GATT), increased from about 10-20 in the early 1980s to over 400 in 1999. Low- and middle-income countries reported that, from 1996 to 1999, they were unable to meet sanitary and phytosanitary requirements on more than 50 percent of their potential exports of fresh and processed fish, meat, fruit, and vegetables into the EU (OECD, 2001a). In fact, these measures were viewed as more important barriers than tariffs and quotas. The WTO Agreement on Technical Barriers to Trade and Agreement on the Application of Sanitary and Phytosanitary Measures are intended to strengthen international rules governing product standards to minimize their abuse, but they entail substantial costs.

Trade preferences. Most developing countries have preferential access to industrial country markets. This departure from the traditional nondiscrimination principle of the GATT has been sanctioned under the Generalized System of Preferences (GSP). In 2001, some 15 such schemes were in effect.

The benefits of many GSP schemes have been limited, however. Although they lower tariffs for exports from many low-income countries, they also divert trade from other countries that are often just as poor. In addition, preference margins are usually smaller for sensitive products, which enjoy the greatest protection, and the fact that a large number of countries—often with similar export structures—benefit from preference schemes reduces any competitive advantage the schemes might convey. (EU preferential trade now covers all but nine countries, though these account for over 40 percent of EU imports.) Preferential tariffs may be costly for their beneficiaries because they are contingent on the imposition and monitoring of rules of origin or of social and environmental conditions. Moreover, evidence suggests that GSP schemes may undermine incentives to engage in trade liberalization and thus perpetuate antiexport biases.

However, recent initiatives like the EU's Everything But Arms initiative—which grants duty- and quota-free market access to LDCs for all products except arms and ammunition (bananas, rice, and sugar will be liberalized gradually)—seem to offer the prospect of additional benefit at a fairly low cost in terms of trade diversion, because LDCs account for a mere 0.5 percent of world trade (provided that the rules of origin, which determine how much value added has to come from the exporting country, are liberal). Simulation results suggest that, if such conditions were offered by all Quad countries, exports from LDCs might increase by $2.5 billion, or about 11 percent (Hoekman and others, 2001). The United States has taken one step in this direction with the African Growth and Opportunity Act (AGOA) of 2000, which nonetheless maintains a number of restrictions, such as tight rules of origin and other conditions for eligibility that many African LDCs have found difficult to meet.

Preference schemes are not a long-term solution. There is a danger they may be seen as a substitute for broader liberalization or even turn into an obstacle to it by creating vested interests in the status quo. They should therefore be set firmly within a context of rapid multilateral liberalization.

Agriculture

Many of the trade barriers described above keep poor countries' agricultural products out of rich country markets. The subsidization of agriculture in the OECD countries depresses world prices of commodities and increases price volatility, which hurts poor countries and their poorest citizens. Agriculture is the dominant economic activity in rural areas, where three-fourths of the world's poor live. It accounts for about 27 percent of GDP in developing countries, a similar share of exports, and 50 percent of employment.

The costs to the global economy of distortions in agricultural trade are large. IMF staff simulations with the Global Trade Analysis Project (GTAP) model (a comparative static, multicountry general-equilibrium model based on neoclassical trade theory) suggest that, even if only static effects are considered, the welfare costs of agricultural distortions may be over $120 billion (based on 1997 data). One-fifth of the cost is borne by developing countries, and the export revenues lost are much larger. Both developing and developed countries suffer the most from their own restrictive policies (see Table 2).

In the OECD countries, total public support for agriculture amounted to $311 billion, or 1.3 percent of GDP, in 2001 (see Table 3). Producer support—domestic subsidies for farmers and border measures (import tariffs and export subsidies)—was estimated to be nearly one-third of total farm receipts. Prices received by OECD farmers were, on average, 31 percent above world prices. Border measures have the greatest distortionary effects on trade. In addition, they are regressive—low-income consumers in OECD countries, who spend a larger share of their income on food, are disproportionately affected, while large farms receive the lion's share of the support. A large share of support is directed at temperate-zone agriculture, but support for products of interest to tropical suppliers is often particularly high as a share of producer receipts. For example, Brazil, a major developing country exporter of food products, faces enormous barriers (see box).

Agricultural subsidies in non-OECD countries are limited compared with those in the OECD countries, but applied tariff levels are similar. However, non-OECD tariff bindings (legal commitments not to raise tariffs over a certain level) under WTO agreements tend to be well above applied rates and applied rates change often, both of which create uncertainty.

While the reduction of tariffs and elimination of subsidies—in combination—in agricultural trade would benefit both developing and industrial countries overall, the distribution of these benefits is more complex and depends partly on the sequencing of liberalization. Certain net food importers might suffer terms of trade and related net income losses initially as prices adjust. They are more likely to experience such losses if subsidy removal is phased in ahead of tariff reductions, because subsidies are highest on products of which developing countries as a group are net importers. Within individual countries, the global liberalization of agriculture would generally benefit developing country farmers, but consumers may face higher prices. Agricultural liberalization will also entail the loss of preference margins for some countries and expose them to tougher competition.

It should be noted that changes in agricultural policies would most likely be phased in very gradually, slowing terms of trade changes and reducing adjustment pressures. Agricultural tariff reductions in developing countries themselves would help mitigate problems by lowering consumer prices, especially for food staples, which tend to be highly protected. In the long run, agricultural liberalization should lead to increased farm investment and enhanced technologies and productivity, and net food importers may become net food exporters. Nevertheless, special attention needs to be paid to food security issues and the concerns of consumers, especially the urban poor. These issues must be addressed comprehensively, as part of national poverty reduction and development strategies; trade policy is unlikely to be the most effective or appropriate instrument for addressing them.

The OECD countries have been sending mixed signals about reforming their agricultural trade policies. Producer support began to decrease in the late 1980s, reaching its lowest level in 1997. Then, as world prices of major commodities fell, it started going up again. In May 2002, the United States introduced a farm bill significantly boosting agricultural subsidies (however, producer support as a share of farm income in the United States is still less than in many other OECD countries, including the EU; it is largest by far in Japan, Korea, Norway, and Switzerland). The recently unveiled mid-term review of the EU's Common Agricultural Policy (CAP) aims to delink farm subsidies from production levels, a move that would curb overproduction, thereby reducing pressure on world prices. The reform proposals are, however, silent on export subsidies and tariffs and must still be approved by member states, a process that could be tricky.

Textiles and clothing

Historically, textiles and clothing have played a unique role in economic development and poverty reduction. Their role in the Industrial Revolution in Western Europe and North America is well known; today, they are spearheading industrialization in the developing world.

From the mid-1960s to 1998, the developing countries' share of world textile exports grew from 15 percent to 50 percent, and of world clothing exports, from less than 25 percent to 70 percent. Total exports of textiles and clothing by developing countries as a group reached $213 billion in 1998. (Sub-Saharan Africa accounted for less than 2 percent of this figure.) Textiles accounted for 51 percent of Pakistan's merchandise exports in 1999, clothing for 50 percent of Sri Lanka's; among the LDCs, textiles and clothing represented 83 percent of Bangladesh's merchandise exports in 1999 and 89 percent of Cambodia's in 2001.

This rapid growth occurred despite high tariffs in both the OECD and the developing countries and extensive quantitative restrictions in the former. OECD tariff peaks affect 27 percent of total tariff lines on textiles and clothing, while the trade-weighted average applied tariff in developing countries is 16 percent. The largest developing country exporters tend to have the highest tariffs. ASEAN (the Association of Southeast Asian Nations), China, and South Asia have tariffs ranging from 20 to 33 percent on textiles, and from 30 to 35 percent on clothing. (Duty exemptions are extensive, however.) These barriers among developing countries are increasingly important, given the growing share of their trade in textiles and clothing with each other.

Quotas for textile and clothing imports in the industrial countries further distort trade in this sector. For nearly half a century, world trade in textiles and clothing has been subject to quantitative restrictions, beginning with Japan's 1955 "voluntary restraints" on its exports of cotton fabrics and clothing to the United States, which evolved in stages into the Multifiber Arrangement (MFA) in 1974. The MFA expanded quantitative restrictions beyond cotton products and was extended several times until the Uruguay Round Agreement on Textiles and Clothing (ATC) took effect at the beginning of 1995. The most competitive exporting countries, like China and India, have faced the most stringent quantitative restrictions, while other suppliers have been able to take advantage of this situation by charging higher prices. Table 4 shows import tariffs on textiles and clothing as well as the export tax equivalents of the MFA quotas (which are administered by the exporting country).

IMF staff simulations with the GTAP model suggest that as many as 19 million jobs for low-skilled workers may have been forgone in developing countries because of MFA quotas, 27 million because of quotas and tariffs combined. Each job saved in a developed country by tariffs and quotas is estimated to cost about 35 jobs in developing countries. In industrial countries, MFA quotas and tariffs hit low-income households, which spend a larger share of their income on necessities, the hardest. (See "Picture This")

According to the same simulations, the combined effect for developing countries of quotas and tariffs on industrial country imports amounts to welfare losses of $24 billion a year and lost export revenues of $40 billion (Table 5). Industrial countries suffer around half the welfare loss and similar export losses. MFA quotas and tariffs also depress demand for products—fiber crops, for example—used in the manufacture of textiles and clothing. With full liberalization, sub-Saharan Africa's cotton exports would be boosted by 9 percent, or about $132 million (at 1997 prices). Extending the GTAP simulation to cover liberalization of textile and clothing imports in both industrial and developing countries, we find that a large share of the developing countries' gains in welfare and exports would come from eliminating their own tariffs.

Although the ATC calls for quota restrictions to be gradually abolished over 1995-2005, Canada, the EU, and the United States have removed very few restrictions. (Norway, the other main user of MFA quotas, has all but eliminated them.) Instead, there has been a "backloading" of liberalization—most of the sectors "liberalized" so far were not restricted in the first place. This might turn what could have been a gradual adjustment into a shock at the end of the transition period—for both importing and exporting countries. A reallocation of production—to the detriment of developing country exporters who have been effectively protected from more competitive suppliers by the quota system—would undermine external balances and could impose high adjustment costs, in view of the large share of textiles and clothing in the exports of a number of countries. With respect to industrial countries, there are also serious concerns that political pressures might spark greater recourse to other forms of protection when quotas are phased out.

The most obvious way to mitigate the shock of adjustment is to accelerate quota removal. Another is to reduce tariffs. Scheduling a gradual path for the multilateral liberalization of tariffs that starts early—in parallel with quota removal—and provides time for adjustment may allow trade negotiators to strike the right balance among the interests of individual countries. Trade liberalization in a broad range of industries would also help offset the loss of market share in the textile and clothing industries.

Going forward

Further liberalization of merchandise trade, especially of agricultural products and textiles and clothing, could generate large benefits for developing countries in terms of incomes, exports, and employment. These benefits would derive not only from the elimination of access barriers to industrial country markets but also from reform of the trade regimes of developing countries themselves. Overall, the opening of markets is a win-win proposition—both industrial and developing countries gain. Many developing countries have been able to develop vigorous and diversified export sectors despite existing hurdles to market access, but better access would no doubt ease the task.

The evidence reviewed in this article (which does not cover trade in services) suggests a number of priorities for moving toward a multilateral trading system that takes special account of the interests of developing countries, including eliminating tariff peaks and escalation; tightening disciplines on recourse to trade remedies; providing more capacity-building assistance to developing countries to enable them to navigate technical and health-related barriers; extending full duty- and quota-free access for exports from the least developed countries; pursuing a comprehensive approach to liberalization in agriculture, including the decoupling of domestic support to agriculture; and accelerating the phaseout of MFA quotas in textiles and clothing trade combined with tariff reductions.

With the benefits of liberalization come a number of risks and adjustment needs. It is important to identify these early and take appropriate action. Liberalization in agricultural trade can have complex distributional effects. When the groups affected are economically vulnerable, supportive policies might be called for; these should, however, be embedded in broader poverty reduction or development strategies. Another risk is that liberalizing quotas in textiles and clothing trade will expose the lack of competitiveness of some developing country exporters. It is crucial for developing countries to assess their competitive position in a post-MFA world and for liberalization to be implemented in ways that minimize its likely impact on the balance of payments and adjustment pressures.

In the context of the Doha Development Agenda, the WTO is committed to negotiations aimed at substantially improving market access for agricultural and industrial products, in particular for developing countries (see "The Doha Development Agenda" in this issue). Increasing market access for developing countries is indeed a necessary first step in helping them grow their economies. But it is not sufficient. It must be part of a broader strategy in the developing countries themselves to promote a vigorous supply response. Inefficiencies in key infrastructure sectors like telecommunications, transport, and financial services often add more to these countries' export costs than foreign trade barriers. And further reform of trade policies and the investment environment in developing countries will be necessary complements to better market access.

References:
  Kym Anderson, Betina Dimaranan, Joseph Francois, Thomas Hertel, Bernard Hoekman, and Will Martin, 2000, "Potential Gains from Trade Reform in the New Millennium," paper presented at the Third Annual Conference on Global Economic Analysis, Monash University, Melbourne, Australia, June.
  Bernard Hoekman, Francis Ng, and Marcelo Olarreaga, 2001, "Eliminating Excessive Tariffs on Exports of Least Developed Countries," World Bank Policy Research Working Paper WPS 2604 (Washington).
  International Monetary Fund, 2001, "Market Access for Developing Countries' Exports," SM/01/137, Revision 1, 8/23/01 (Washington).
  ———, forthcoming, World Economic Outlook, October 2002—Prospects and Policy Issues (Washington).
  Organization for Economic Cooperation and Development, 1995, "Transfer Efficiency of Agricultural Price Support," in Adjustment in OECD Agriculture: Issues and Policy Responses (Paris).
  ———, 2001a, Agricultural Policies in Emerging and Transition Economies: Special Focus on Non-Tariff Measures (Paris).
  ———, 2001b, Market Effects of Crop Support Measures (Paris).
  World Bank, 2002, Global Economic Prospects 2002: Making Trade Work for the World's Poor (Washington).