Clemens F.J. Boonekamp
Successive rounds of multilateral trade negotiations under the aegis of the General Agreement on Tariffs and Trade (GATT) have progressively reduced the importance of tariffs as barriers to trade. This, however, has not been matched by a corresponding liberalization of international trade. Since the early 1970s, with the relatively slow rates of growth and the rise in surplus capacity in many countries, protectionist pressures have steadily built up, especially in industrialized countries whose international competitiveness in traditional industries has declined, in particular vis-à-vis Japan and the newly industrializing countries of Asia. As a result, nontariff barriers have become more prevalent, to the point that they currently rival, if not exceed, tariffs as impediments to trade.
Voluntary export restraints (VERs) are now a common form of nontariff barrier, having grown in number and spread, in recent years, from textiles, clothing, steel, and agriculture to automobiles, electronic products, and machine tools. This article discusses the basic elements of VERs, why they are employed, and their economic consequences.
What are VERs?
A VER is a measure by which the government or an industry in the importing country arranges with the government or the competing industry in the exporting country for a restriction on the volume of the latter’s exports of one or more products. By this definition, the term VER is a generic reference for all bilaterally agreed measures to restrain exports. Strictly speaking, however, a VER is an action unilaterally taken and administered by the exporting country and is “voluntary” in the sense that the country has a formal right to eliminate or modify it. Usually a VER arises because of pressure from an importing country; it can then be thought of as “voluntary” only in the sense that the exporting country may prefer it to alternative trade barriers that the importing country might use. Further, in a noncompetitive, especially oligopolistic, industry exporting firms might find it to their advantage to negotiate a VER, which is then truly “voluntary.”
A VER that consists of a government-to-government agreement is normally referred to as an orderly marketing arrangement and often specifies export management rules, consultation rights, and the monitoring of trade flows. In some countries, notably the United States, orderly marketing arrangements are legally distinct from a VER as strictly defined. Agreements which involve industry participation are often referred to as voluntary restraint arrangements. The distinction between these forms of VER is largely legal and terminological and has little bearing on the economic impact of VERs.
A typical VER limits the supply of exports, by commodity type, by country, and by volume. GATT Articles, which are concerned with governmental actions affecting trade, prohibit export restraints under normal circumstances; when permitted, they must be nondiscriminatory and applied only through duties, taxes, and charges. However, the involvement of governments in VERs is not always clear. Also, VERs do not always have firm market-sharing provisions; they may, for example, be in the form of an export forecast and thus become cautionary in nature. For these reasons, VERs fall into a “grey area” in that there can be doubt as to their illegality under GATT. Further, parties to a VER are unlikely to request a finding under the GATT’s dispute settlement procedures—they have never done so—while third parties can often appear to benefit from a VER and may therefore be reluctant to initiate dispute procedures. Finally, signatories to the Code on Subsidies and Countervailing Duties Code that resulted from the Tokyo Round of trade negotiations seem to have acquired legal powers to negotiate VERs. It is in these respects that the decision of the GATT Council of Representatives in 1987 to establish a dispute panel to examine the Japan-US semiconductor agreement is important.
Why are VERs introduced?
Broadly speaking, restrictive trade measures are taken usually for two purposes: to protect or improve the balance of payments situation, and to provide relief for industries adversely affected by foreign competition, in principle allowing them time to undertake the adjustments necessary to regain external competitiveness. VERs are employed for this latter purpose and, compared to other forms of protectionism, offer several advantages from the viewpoint of the protecting country.
Under GATT rules, safeguard provisions exist for the temporary, emergency protection of domestic industries injured by import competition. Such safeguard actions, however, involve negotiating compensation with the countries affected by the measures. These negotiations can be difficult and may not succeed. In that case, the protecting country risks retaliation by the exporting country. In either case—compensation or retaliation—the exports of the country taking the safeguard action may suffer. VERs, on the other hand, have built-in compensation, in the form of rents (i.e., higher earnings arising out of the scarcity of a product). This makes acceptance by the exporting source more likely and retaliation less probable.
The importing country, in negotiating a VER, tends to avoid the frequently lengthy, public, and often multilateral, debate that invariably precedes other forms of protectionism, such as increasing tariffs or imposing quotas. In such a debate, the cost of the protective measure is likely to become more clearly recognized, making the action politically expensive and risky. A VER then has the advantage that, as an action taken by a foreign source, a domestic legislative struggle may be avoided; it often can be negotiated quickly without its costs becoming obvious. Further, in the context of exports which are, or are suspected of being, subsidized, the domestic authorities can bypass the often expensive and time-consuming process of a countervailing-duty investigation by coming to a VER agreement with the exporter. Finally, it can be argued that a VER, by addressing the source of the problem, that is the one or few low-cost suppliers disrupting the domestic industry, obviates the need for wider action that could harm third countries, as would be the case with a nondiscriminatory import quota of equivalent import-reducing effect (see below). For any of these reasons domestic policymakers often prefer a VER to alternative measures; it offers relatively quick, politically inexpensive assistance to an industry threatened by import competition.
A VER can also be attractive to the exporting country. As indicated above, it offers rents which, at least in the short-run, are windfall gains to the extent that demand in the rest of the world is elastic so that terms of trade losses are minimal or zero. This can be important relative to alternative trade barriers that the importing country might take; tariff revenues, for example, accrue to the government that levies them. Also, VERs can serve to assure the exporter of market access to the importing country, terminate the uncertainty inherent in a countervailing-duty investigation, and provide the exporting government with an element of control over its domestic industry. These factors suggest that when faced with the probability of protection by the importing country, particularly if it is an important market for other products, the exporter might agree readily to a VER.
Prevalence of VERs
Quantitative export restraints first seem to have appeared in 1935 when Japan was induced to limit its textiles exports to the United States. However, only in the past decade or so have they come into widespread use. The accompanying table lists almost 100 major, known VERs. The actual number may well be greater as there are reported to be various undisclosed industry-to-industry and govemment-to-industry arrangements. Of the known number of VERs, 55 restrict exports to the European Community or its individual member states, and 32 restrain exports to the United States. In general, VERs have been introduced to protect industries in OECD markets where certain developing countries, East European countries, or Japan have become serious competitors. Indeed, the exports of these countries are restrained by about 80 VERs.
The above numbers do not include the Multi-Fibre Arrangement (MFA), which together with its predecessor, the Long-Term Arrangement on Cotton Textiles (1962-72), has been the model for VERs. The MFA is a negotiated, multilateral departure from GATT based on the principle that industrial countries, which are the main importers of textiles, require special protection against “market disruption” by lower-cost, normally developing-country, exports. Under its aegis a multiplicity of bilateral export restraint agreements have been concluded, covering approximately 50 percent of trade in textiles and clothing. In addition, as the table notes, there are 11 known VERs in this sector outside the scope of the MFA. As a result a considerable portion of world trade in textiles and clothing is managed, and thus not subject to the normal forces of international trade.
Apart from textiles and clothing, steel is the product category most heavily affected by VERs. Since the first restraint in this sector was negotiated in 1968 between the United States and several European and Japanese exporters, about a quarter of the total trade in steel has come under VERs, affecting exports from nearly all the major third country suppliers to the United States and the European Community, as well as exports from the European Community to the United States. Exports of agricultural products are also restrained by VERs, principally from the more efficient producers, such as Australia and Argentina, to the European Community. In automobiles and transport equipment, as well as in electronic products and machine tools, Japanese exporters limit their sales to both the European Community and the United States, while in footwear a number of OECD markets are protected by VERs with Korean exporters.
According to one calculation, in 1984 some 10 percent of total world trade, and 12 percent of non-fuel trade, was covered by VERs (“Export Restraint Arrangements” by M. Kostecki, World Economy, forthcoming). This study also estimated that in the same year approximately 38 percent of Japanese exports to the European Community and 32 percent of Japanese exports to the United States were covered by VERs. Other commentators estimate that in 1983 about 11 percent of world trade in developing countries’ manufactured goods was restrained by VERs. Further, the percentages appear to have risen rapidly in the early 1980s. By one estimate the share of exports of the newly industrialized countries of Asia and Japan affected by VERs rose from some 15 percent in 1980 to about 32 percent in 1983.
The economics of VERs
In general, the effect of VERs is to reduce the level of imports and thus increase the price of the product in question in the importing country. This will happen as the normally low-cost, but now restricted, foreign suppliers raise their export prices to capture the rents created by the VER. The higher price will normally encourage an increase in domestic output of the product.
However, even if the VER is set at the free trade level of imports, in an oligopolistic industry prices can rise because a VER fundamentally changes the nature of competition in the affected industry. Given that the exporting industry (or country) administers the VER, producers in the protecting country become “price leaders” relative to those in the exporting country: any increase in the price by domestic producers must result in an appropriate rise in the price of the imported substitute if the VER is to be observed. In these circumstances the profits of both the domestic and foreign firms increase. Consumers will, of course, lose. Producers in the exporting country can be expected to agree willingly to a proposed VER, particularly since it can increase their share of the importing country’s market, if the demand for the domestic product is more responsive to changes in the price of the home product than to changes in the price of the imported substitute. The VER then protects domestic producers, by raising their profits, but does not protect domestic production.
These results become less tenable the larger the number of firms in the industry. First, it becomes more difficult to persuade all domestic firms to be part of the price-leadership role; some may prefer to increase market share by price competition. Second, if not all foreign firms are covered by a VER, they, too, may seek a larger market share. Thus the larger the number of firms in the industry, the more likely that the VER will need to be set at an import-reducing level if the domestic price is to rise and encourage higher local production levels. However, the oligopolistic example does point to the important conclusion that, regardless of the number of firms in the industry, a VER provides an incentive for collusion between firms, to change the nature of competition. This creates vested interests in both the importing and exporting country. A VER may encourage, therefore, what anti-monopoly legislation is meant to forestall.
Price rises due to VERs can be substantial. In a recent study of the impact of the VER on Japan’s automobile exports to the United States (“The Cost of Trade Restraints” by Charles Collyns and Steven Dunaway, IMF Staff Papers, March 1987) it was estimated that the average price of a car sold in the United States in 1984 was about $1,650, or 17 percent, higher than it would have been in the absence of the VER. Of this estimated increase about $1,030 reflected improvements in quality. Exporters under a volume constraint have an incentive to upgrade the quality of their product on the restricted market in order to maximize profit per unit sold. Unrestricted foreign suppliers and domestic producers are likely to follow suit if their products are relatively close substitutes for the improved, export-restrained, product. In the above study the pure price effect was some $620 per car in 1984; this cost US consumers a total of $6.6 billion.
VERs can distort trade and production in a number of ways. First, in the importing country the market share of exporters not subject to VERs can increase. The price of the product subject to the VER in the importing country is likely to exceed the world price; the latter may in fact fall, particularly if the importing country’s market is large and the VER well enforced. In these circumstances, non-restrained suppliers may divert exports to the protected market, thus distorting the pattern of trade. At the time of the 1977 orderly marketing arrangement for color television sets between Japan and the United States, Japan accounted for some 90 percent of US imports of such sets. Two years later Japan’s share had fallen to about 50 percent, while that of the Republic of Korea and Taiwan, Province of China had grown significantly. Subsequently the OMA was extended to Korea and Taiwan.
Second, there may be an incentive for new firms in both the importing and other countries to enter the affected industry, making investments that might not have been efficient at the free trade level of imports. Third, the existing VER-affected firms in the exporting country may seek to circumvent their constraints by exporting via third countries. Finally, foreign suppliers may take to exporting, or increasing exports of, close substitutes for the VER-protected products to the importing market.
If VERs spread so as to cover imports of a product from all sources, the VER-system then resembles an import quota. The process, however, is quite different. A quota is normally applied on a global basis; it is nondiscriminatory, usually being allocated either on a “first-come, first-served” basis or by quota-shares in accord with the previous pattern of import shares. VERs are bilaterally negotiated, normally with one or a few suppliers. They are discriminatory, therefore, with export volumes depending on negotiating strength. They can bias the pattern of trade for the VER-covered product to the importing country against the more efficient exporters and may create investment signals for producers in third countries that could turn out to be wrong. As such, VERs can lead to larger efficiency losses than a globally applied quota of equivalent import-reducing volume.
|Major known VERs|
(excluding the MFA)1
(by number of arrangements)2
(by number of arrangements)2
|Total:||99||Japan (24), Korea (14), Brazil (4);||EC (55)3; USA (32);|
|ODCs (21); OICs (20); E. Eur. (16)||(Canada, Japan, Norway (12)|
|Steel||39||EC (4); OICs (12); DCs (12);||USA (25): EC (14)|
|E. Eur. (11)|
|products||17||DCs (6); ICs (6); E. Eur. (5)||EC (16); Canada (1)|
|Automobiles and transport equipment||13||Japan (11); Korea (2)||EC (9); USA (1); OICs (3)|
|clothing||11||Korea (2); ODCs (9)||USA (4); EC (3); OICs (4)|
|products||7||Japan (6); Korea (1)||EC (6); USA (1)|
|Footwear||5||Korea (3); Japan (1); Taiwan (1)||EC (2); OICs (3)|
|Machine tools||3||Japan (3)||EC (2); USA (1)|
|Other||0||Korea (3); ICs (1)||EX (3); Norway (1)|
As already indicated, VERs can—and have—spread from country to country and from product to product. In this respect their demonstration effect can be important. Moreover, as previously noted, a VER tends to lower the world price of the product, increasing demand in third countries, which might be satisfied by imports of the product. This effect could be reinforced to the extent that the VER-covered exporters are able to use their rents from the protected market to win market share in nonprotected countries. These factors may help to persuade third countries, especially those that manufacture the product in question, also to negotiate VERs. Eventually, as in textiles and clothing, and steel, a VER network could effectively result in a globally managed market-sharing arrangement.
VERs tend to create rents, which can be substantial, for foreign producers. The OECD has estimated that the annual transfer from OECD countries to exporters of textiles and clothing in the newly industrialized countries of Asia is at least $2 billion under MFA bilateral export restraint arrangements. In the above-noted study of the VER on Japan’s auto exports to the United States, it was calculated that Japanese exporters, on pure price effects alone, “earned” rents of $1 billion in 1984; the total transfer to foreign suppliers was $1.67 billion, indicating that third-country exporters not restrained by the VER may have benefited from it. M. Kostecki, using a VER tariff-equivalent method, has estimated that the rent transfers resulting from VERs worldwide in 1984 could have been as much as $27 billion.
Trade policymakers are not unaware of the effects discussed above, in particular of the efficiency losses in the allocation of resources. This is one reason why VERs are generally negotiated to have a relatively short lifespan. Nevertheless, VERs can be a tempting form of protection. Once they are in place, vested interests may work against their removal.
Do VERs work?
The answer to this question depends on whether a VER serves its purpose of safeguarding employment and promoting adjustment in the protected sector, and, if so, at what price. These aims might not be consistent with each other. Modernization to regain competitiveness often entails switching to a more capital-intensive mode of production.
Similarly, if a VER implies a political constraint to retain employment, adjustment by the industry might be delayed. This might happen in any event as the VER, akin to other forms of protectionism, reduces competitive pressures and makes it easier to continue production with outdated and inefficient technology. The industry then retains low-quality staff, displacing potentially higher-skilled employees who could make a greater contribution to raising real incomes.
More generally, by raising domestic profits VERs can make resources available for adjustment. This appears to have happened in the US automobile industry and in certain parts of the textile and steel industries in both the European. Community and the United States. However, in these sectors the VER protection has also afforded the exporters an opportunity to upgrade the quality of their products. The domestic industries thus face increased competition in precisely those areas where they might otherwise have had a comparative advantage.
A recent GATT report shows that VERs have not prevented a loss of employment in the textiles, clothing, and steel industries in the protecting countries. In the period 1973-84, employment in the steel sectors of the European Community and the United States declined by 42 percent and 54 percent, respectively; in textiles and clothing the declines were 46 percent and 43 percent, respectively, in the European Community, and 22 percent and 18 percent, respectively, in the United States. Modernization, gains in productivity, and changes in macroeconomic conditions dominated the employment situation in these sectors, as they did in the US automobile industry where employment fell by some 250,000 jobs in the early 1980s. It is possible, however, that employment in these sectors would have fallen further in the absence of protective measures. Thus, by one estimate between 40,000–75,000 jobs were saved in the US automobile industry during 1981-84 as a result of the VER on Japan’s car exports to the United States. Similarly it has been calculated that a 50 percent relaxation in the VERs protecting the Swedish clothing industry would reduce clothing employment in Sweden by about 6 percent.
Job maintenance by VERs can be very costly. Estimates for 1984 indicate that the annual cost, as a result of higher prices, to US consumers under VERs on textiles and clothing amounted to $50,000 and $39,000, respectively, per position saved, as compared with annual average textile and clothing wages of $13,400 and $10,500, respectively. The OECD has calculated that each job protected under the orderly marketing arrangement on Japan’s color television exports to the United States cost about $60,000 per year.
The prevalence of VERs and their inherent drawbacks, especially their tendency to spread and fragment the trading system into a series of market-sharing arrangements dominated by the major trading nations, has led some commentators to suggest that they should be regulated and their use be subjected to strict conditions and multilateral surveillance. This would encourage greater adherence to internationally agreed trade rules. These observers also note that GATT already sanctions certain departures from the principle of nondiscrimination as, for example, in its provisions for customs unions and free trade areas. Others suggest that the principle of nondiscrimination is too vital to be further weakened by giving VERs official approval. Nondiscrimination in economic terms means that a given level of protection for domestic producers can be achieved at minimum cost to domestic consumers and the rest of the world. It also protects the interests of smaller trading nations and helps to ensure the access of new entrants to the international market place. Its role in GATT rules lends transparency and predictability to international trade relations and to domestic decisionmaking. In this view, VERs would be controlled by a firmer commitment by trading nations to nondiscrimination in the formulation and practice of their trade policies.
This issue is on the agenda of the recently launched Uruguay Round of multilateral trade negotiations, especially in the context of GATT’s safeguard provisions. The decisions of the negotiators could have an important effect on the future of the international trading system and by implication on the growth and employment prospects of its member countries.