Anti-dumping Policy: An Overview of the Research

Nigel Grimwade[1]

Paper Number 1-09

Research Working Papers


Over the last thirty years, interest in anti-dumping has grown as the number of

countries involved in anti-dumping has increased. This has been reflected in

a huge increase in the amount of research undertaken on antidumping and in the

volume of research published on the topic. This paper seeks to provide a brief

summary of the work done and the conclusions that can be drawn from it. It

concludes with some pointers to new avenues of research.

The paper reviews four areas of research on anti-dumping–namely, the theoryof dumping, themeasurement of anti-dumping activity, the determinants of antidumping and the effects. The basic theory of dumping is now pretty wellestablished, although important new work has been done on the effects ofdemand uncertainty and wage rigidity on the occurrence of dumping.

A lot of work has alsobeen done to refine the measurement of anti-dumping activity and to gage its intensity over time using the extensive data that is now available on antidumping. However, it remains unclear whether or not the degree of activity is on a rising trend.

Econometric studies of the determinants of anti-dumping point toa complex set of micro- and macro-economic factors that affect the trendof anti-dumping activity over time. There is some suggestion that thefactors determining the use of anti-dumping may not be the same for developedas for developing countries.

Finally, analytical studies of the effects of anti-dumping generally support theconcerns that anti-dumping has harmful effects on competition and distortpatterns of trade. Changes to the WTO anti-dumping code such as the introduction of a sunset clause havedone something to limit these effects.However, the case for further revisions to the code remains a strong one.

Key words: Anti-dumping, dumping, trade, trade liberalization, competition, economic welfare

Anti-dumping Policy: An Overview of the Research

1. Introduction

A deep interest among academic economists and other researchers in the phenomenon of anti-dumping has only really developed in the last thirty years. Yet, anti-dumping has a long history. Writers going back as far as the 18th century refer to cases of underselling by foreign producers. Then, as now, such practices were viewed as harmful and attempts were made to deal with them. Despite this, it was not until the early 20th century that legislation was passed to counteract dumping. The first anti-dumping laws were introduced by Canada in 1904, followed in 1905 and 1906by New Zealand andAustralia respectively and by the United Statesin 1916. In 1947, the right of countries to take action against dumping that caused or threatened material injury to domestic producers was enshrined in Article VI of the General Agreement on Tariffs and Trade (GATT). In 1968, the EC formally introduced an EU-wide anti-dumping policy as part of the EC’s common commercial policy.

However, although anti-dumping cases were not uncommon in the early post-war era, the reported cases of anti-dumping usage appear to have become significant only after 1980. At this point, anti-dumping policy began to attract lively interest from trade economists and other researchers and the number of papers in journals on the subject accelerated. This paper seeks to provide an overview of the results of the published research on anti-dumping. The purpose is to set out what we have learnt about anti-dumping as a result of the research undertaken and to set out some pointers for future research. We begin with a brief outline of anti-dumping rules and their application. This is followed by a discussion of each of the main avenues of research undertaken and the results obtained. The paper concludes with an attempt to identify current gaps in the knowledge about anti-dumping with some suggestions as to where further work is required.

2. What is anti-dumping policy?

Under Article VI of the GATT (1969), dumping occurs “if the products of one country are introduced into the commerce of another country at less than the normal value of the products”. Article VI:I states that normal value is given by either:-

(a) “the comparable price, in the ordinary course of trade, for the like product when destined for consumption in the exporting country, or…

(b)” in the absence of such domestic price:-

(i) the higher comparable price for the like product for export to any third country in the ordinary course of trade, or

(ii) the cost of production of the product in the country of origin plus a reasonable addition for selling costs and profits”

(Article VI:1, GATT)

The latter two provisions are important as they allow for the possibility of comparing the export price with either the price of a like product in a third country or a constructed price based on costs in the importing country plus a “reasonable margin” for profit.

It is important to note that, whatever the method used for determining the normal value, the comparison must be with a “like product” and must take place at the same time. A further requirement is that the comparison must take place “at an equivalent level of trade”. This is normally taken to be the ex factory price of the product i.e. the price of the good when it leaves the factory gates and, therefore, not including all costs incurred subsequently, such as costs of distribution, distributor’s profit margin or any taxes imposed. This necessitates making deductions from the final selling price of the good in the exporting and importing countries before the comparison is made. This has become a source of contention because the determination of dumping hinges greatly on which costs are deductible and whether the same deductions are made for domestic sales as for exports.[2]The difference between the export price and the normal value is known as the “margin of dumping”. Article VI states that, where dumping is found to have taken place and to have “caused or threatened material injury to an established industry” or “materially retarded the establishment of a domestic industry”, countries may impose anti-dumping duties on the exporters found to have engaged in harmful dumping. However, the rate of anti-dumping must not exceed the estimated margin of dumping. Some countries (notably, the US) tend to set anti-dumping duties right up to the dumping margin; others set duties somewhat lower (e.g. the EU).

In the 1960s, as part of the Kennedy Round of multilateral trade negotiations, an Anti-Dumping Code was agreed to govern the use of anti-dumping policy. This has subsequently been amended and added to in each subsequent GATT round. The current code was substantially revised as part of the Uruguay Round which was concluded in 1994 and is currently the subject of negotiations within the framework of the Doha Round. Members of the WTO are permitted to introduce their own laws for combating dumping and these laws must conform to the WTO Code for anti-dumping. Should this not be the case, another member state may bring a complaint before the WTO, setting in motion an adjudication process within the WTO[3]. The Code makes clear that, before anti-dumping measures are imposed, the authorities of the importing country must carry out a formal investigation of the complaint. As part of the investigation, all interested parties, including the exporters, are allowed to submit any information relevant to the case. Following a preliminary finding, provisional duties may be imposed for a period of between four to six months. However, during this period, a fuller investigation has to take place. In the light of this, the duties may be removed or replaced with definitive duties, lasting up to five years. A somewhat disturbing trend has been for anti-dumping cases to be settled through exporters making price undertakings, rather than through the imposition of actual duties.[4]

Key parts of the WTO Code deal with the procedure for calculating the normal value of a product, the procedure for estimating the margin of dumping, the determination of “material injury” and the imposition of anti-dumping measures. For example, there are special provisions for where the importing country is a non-market economy. As prices in such a country are state-determined, the margin of dumping cannot be determined by comparing the export price with the price in the exporting country. Instead, the importing country is permitted to use a surrogate domestic price, using domestic costs of production for a like product in some third country with the usual addition for selling costs and for profit. Perhaps, not surprisingly, WTO rules have been much criticised for giving too much freedom to the importing country to ensure an affirmative outcome. Rules for the determination of injury have also attracted controversy. These set out the criteria to be used in for determining whether or not dumping has resulted in or threatens injury, such as falling sales, declining market share, falling output, falling prices, declining profits, under-utilisation of capacity and so forth. However, what is less clear is whether such factors are to be taken as the result of dumping or whether they are due to other developments, such as a loss of competitiveness.

One key issue in the Uruguay Round negotiations on anti-dumping was the issue of circumvention. Both the United States and the EC (now EU) had introduced provisions for imposing anti-dumping duties on foreign producers who set up so-called screwdriver plants as a way of avoiding such measures. A screwdriver plant is an assembly operation inside an importing country which is created for the sole purpose of avoiding duties. It does so by importing components and parts or semi-assembled goods, which are not subject to any duty, and then selling the finished good in the importing country with impunity. In most cases, these measures required some proportion of the output to be produced from within the importing country. Such measures were viewed by some countries as a means of forcing foreign producers to invest in the importing country, thereby benefiting the economy of the importing country at the expense of that of the exporting. Both the US and the EC wanted WTO rules to be amended to permit such measures. They were unsuccessful.

3. Why does anti-dumping policy matter?

As anti-dumping policy has played a more important role in trade policy, so it has attracted growing interest from both academics and policy-makers alike. These have included both economists and lawyers. Important issues to be address include the following. Whether or not dumping is harmful and, therefore, whether or not anti-dumping laws are needed. What constitutes “fair” trade and whether antidumping laws prevent fair competition from taking place. Whether the spread of anti-dumping is a form of backdoor protectionism that is gradually taking the place of more traditional forms of protectionism and threatening the expansion of world trade that has taken place over the past half century.

One reason for these concerns is the evidence for a big increase in both the extent and intensity of anti-dumping activity. Here the evidence is not compelling. The number of anti-dumping cases has actually been highly volatile, falling in some years and rising in others. This suggests that anti-dumping may be highly affected by the particular stage reached in the business cycle. For example, there was a sharp increase in anti-dumping activity in the first six months of last year. However, the number of cases initiated in a particular year is a poor measure of the intensity of anti-dumping activity, as it takes no account of the proportion of world trade that is affected by such measures. Nor does it take into account the outcome of cases initiated, nor how long the measures last for. Also important is the level of duty imposed and the number of exporters affected.

A particular concern must be theincrease in the number of countries with anti-dumping laws, which are now making active use of these laws. In the 1980s, the main users of anti-dumping were the developed countries, specifically, the United States, EC, Australia and Canada. Between 1980 and 1984, 97% of anti-dumping complaints were filed by these four countries (Prusa, 2005). However, beginning in the mid-1980s, there has been a spread of anti-dumping to a wide range of other countries. These include a number of developing countries, most notably, India, Argentina, Mexico and South Africa. In recent decades, these new users have come to account for all of the increase in reported anti-dumping activity. Of course, to a large extent, this is due simply to the growing role played by emerging nations in world trade. However, it should also be seen as a response by these countries to the fact that developed countries have made so much use of anti-dumping and that this is often viewed by the countries targeted as a form of protectionism through the backdoor. In short, the spread of anti-dumping to new users is a form of import retaliation, which could pose a threat to open markets and trade expansion if not addressed.

In the past, the main advocates of anti-dumping policy were without doubt the developed countries. Anti-dumping was widely regarded as a necessary response to a situation where imports are being sold at a price that is below the costs of production or “less than fair value”. It was argued that countries should defend themselves against such unfair practices. Over time, however, this traditional defence of anti-dumping has been devalued as a result of the growing misuse of the policy. As we have seen, anti-dumping laws permit anti-dumping measures to be imposed whenever the export price is below the “normal value” of the product. Thus, the definition of harmful dumping has been extended to cover any situation where the export price is below the domestic price not simply below costs of production. Moreover, because many anti-dumping cases are settled using the price in a third country or even a constructed price, there can be no certainty that even this is true. This has led many economists to question whether, in fact, dumping, even as defined by the GATT, is actually taking place in the majority of cases. Indeed, the margins of dumping found in many of the cases actually reported are so high as to beggar belief. The result has been the imposition of anti-dumping duties well above the level of most-favoured nation MFN tariffs applied anywhere in the world.[5]

When anti-dumping came into greater use, the targeted countries were mainly other developed countries. Thus, most anti-dumping measures were imposed by developed countries on imports from other developed countries. However, beginning in the late 1980s, this situation gradually changed. Many anti-dumping measures were targeted against emerging nations in Asia. Today, the East Asian countries – especially, China, South Korea, Japan, Taiwan and Thailand – are the main victims (Prusa, 2006). At the same time, as we have seen, emerging nations have become the major users of anti-dumping. This is likely to change the traditional stance of countries such as the US and EU to negotiations to revise the WTO anti-dumping code.

4. Theoretical studies: the theory of dumping

Economic research on anti-dumping has followed a number of paths. First, some research has been concerned with exploring the reasons why firms engage in dumping and the effects of dumping on economic welfare. Pioneering work in this field was carried out by Jacob Viner in 1923, who explained dumping using the familiar model of price discrimination where markets are geographically separated (Viner, 1923). Dumping is viewed as simply another form of price discrimination, except that the two markets are separate countries. Thus, dumping is the result of the exercise of monopoly power by a single supplier plus the ability to segment markets geographically to prevent re-exporting. The rationale for such discrimination was the existence of different elasticities in different markets for the same product, which creates the potential for the exporter to maximise profits by dividing his output between the two markets so as to maximise his total revenues. Rather than reducing the economic welfare of the importing country, however, such dumping increases welfare by reducing the price of the good in the importing country and raising the real incomes of consumers.

Students of industrial economics, however, will be more familiar with the model of predatory pricing, in which a dominant firm with superior financial resources drives competitors out of the market by setting price below average costs. Having eliminated all rivals, the predator then raises price above marginal costs and recovers his losses. This is often thought to be the motive behind dumping. However, it has since been demonstrated that cases of predatory dumping are comparatively rare.[6] This is because the conditions required for predatory pricing to be an optimal strategy are highly restrictive. Clearly, however, such dumping is harmful, as reduces competition in the market of the importing country and leads to higher prices in the long run.