United States General Accounting Office GAO Report to Congressional Requesters November 1990 INTERNATIONAL TRADE Comparison of U.S. And Foreign Antidumping Practices GAO/NSLAD-91-59 National Security and International Affairs Division B-240122 November 7,199O The Honorable Bob Packwood Ranking Minority Member Committee on Finance United States Senate The Honorable Daniel Patrick Moynihan United States Senate As requested, we are providing a comparison of the antidumping policies and practices of the United States and its major trading partners. This report follows our July 1990 fact sheet on the use of the GATT Antidumping Code(GAO/NSWW~~SFS). We are sending copiesof this report to the Secretary of Commerce,the Office of the U.S. Trade Representative, and the International Trade Commission.Copieswill also be made available to other parties upon request. Pleasecontact me on (202) 275-4812 if you or your staff have any questions concerning this report. The major contributors to this report are listed in appendix II. Allan I. Mendelowitz, Director International Trade, Energy, and Finance Issues Executive Summary During the current Uruguay Round of the General Agreement on Tariffs and Trade negotiations, somesignatories to its 1979 Antidumping Code are seeking to limit the use of antidumping measuresand to more clearly define what constitutes dumping and its injurious effect on domestic industries. Senators Daniel Patrick Moynihan and Rob Packwood asked GAO to com- pare the antidumping policies and practices of certain major trading partners with those of the United States, focusing primarily on the fol- lowing areas: (1) how much evidenceis neededto initiate investigations, (2) how open, or “transparent,” antidumping practices are to the parties involved and to the public, and (3) what rights of appeal on antidumping decisions are available. The Antidumping Code establishesguidelines for measuresto counter Background the ill effects of dumping-defined as the sale of exported products at a price lower than that charged for the sameor a like product in the home market of the exporter. To initiate an investigation, the code requires that signatories have “sufficient” evidenceof dumping and resulting injury and that petitions be made by or on behalf of the industry I affected. The code also calls for using procedures that will result in an equitable and open antidumping process.The code sets guidelines for the use of antidumping measuresand related practices, but member countries are responsible for implementing the codeunder their own laws and regulations. GAO examined the practices of traditional users of antidumping laws (Australia, the United States, Canada,and the European Community) and a new user, Mexico. Other US. trading partners have had little prac- tical experience with antidumping actions. Sufficiency of evidence is not fully defined by the Antidumping Code. IZesultsinBrief As a result, signatories can exercise a great deal of discretion in reviewing antidumping petitions and their supporting evidence when determining whether formal proceedingsare necessary.Industry repre- sentatives and trade specialists have expressedspecific concernsthat the level of evidence required to initiate an investigation is low in the United States, the European Community, and Mexico. Fewer complaints were directed at Canada and Australia. P8ge2 Execntlve snlnmuy Although the code doesnot explicitly require that a party have “standing” to bring a petition (which refers to determining whether petitioning parties represent a major proportion of their industry), industry representatives and trade specialists have also expressedcon- cern about the perceived laxness in practices for determining standing. The degreeof transparency varies among the signatories reviewed. In general, the United States appears to have the most open antidumping system, while the Mexican and European Community systems seemthe least open in providing information on their decision-making process. Each of the signatories included in GAO'S review provides appeal rights to affected parties through administrative and judicial review. Signato- ries’ administrative review of antidumping determinations differs in impact, frequency of use, and practice. Whether or not antidumping duties automatically terminate after a certain period of time greatly affects the importance and use of administrative review. The scopeof judicial review is broader in the United States than in the other signatories. GAO’sAnalysis Sufficiency of Evidence Trade experts and representatives of industries accusedof dumping have raised concernsabout the large amount of discretion signatories have in deciding whether evidenceto initiate antidumping investigations is sufficient. The decision to initiate is important becauseof the poten- tially disruptive effects antidumping investigations can have on trade. U.S. initial reviews of antidumping allegations and supporting evidence are viewed by many as less rigorous than those of other signatories pri- marily because(1) the Department of Commerceis precluded from con- sidering information from respondentsat the initial stage;(2) the U.S. antidumping system fosters the initiation of investigations by focusing almost exclusively on dumping, rather than on injury and causation; (3) Commercestrictly adheresto set time frames; and (4) Commerce assumesthat the petitioner has standing unless proven otherwise. GAO'S review indicated that only 6 out of an estimated 171 petitions received by Commercefrom 1986 through 1989 were dismissed without Executive Summary investigation. However, a Commerceofficial explained that many poten- tial petitions are eliminated through informal discussions. The European Community and Mexican initial reviews are also thought to be less rigorous, although for different reasons.For example, GAO'S review indicated that the Community’s screeningprocessfocused prlma- rily on the determination of injury and was less strict regarding evi- denceof dumping. The laxness with which signatories determine whether a petitioner has “standing” has raised concernsthat petitioners without standing may causeinitiation of an investigation without sufficient evidenceof i@uy, Such weakly supported investigations place a heavy administrative burden on the accusedparties and can disrupt trade. Transparency Although the transparency of procedures followed by all major signato- ries has increased appreciably in recent years, U.S. procedures are still widely believed to be more transparent than those of Australia, Canada, the European Community, and Mexico. This opennessis a function of several factors, including U.S. public hearings, its comprehensiveand publicly available antidumping decisions,and its thorough disclosuresof the basesof its decisions.Australia and Canada also have fairly trans- parent procedures, although both lack someof the features of the U.S. system. Conversely, antidumping participants have criticized the procedures of the European Community and Mexico for their lack of transparency, although they have noted that procedures have improved in the Euro pean Community. Rights of Appeal While the Antidumping Code doesnot provide for appeal rights, admin- istrative and judicial review is available in the signatories GAO examined. Administrative review generally involves modifying or revoking antidumping measures,basedon changedprices or conditions. , GAO found that the use of such review differs among the signatories. For example, the United States performs a large number of administrative reviews becauseit ‘relies on these reviews to determine actual dumping amounts. In addition, affected parties use administrative review in attempting to end the imposition of antidumping duties. These actions Pyre4 Esecutlve Summary occur becauseUS. law, as well as Mexican law, lacks a “sunset provi- sion,” which automatically terminates antidumping duties after a fixed period of time. Judicial review, which involves appeals of antidumping decisions made to courts, is rather limited in most countries becausethe scopeof the courts review is primarily confined to issuesof law. However, the U.S. Court of International Trade has interpreted its scopeand jurisdiction broadly, to include review of factual issuesnot strictly related to the law. As a result, judicial review in the United States is thought to be more extensive and effective than that of others included in GAO'S review. Affected parties are more likely to bring appeals in the United States than elsewhere. Parties in Australia, Canada,the European Com- munity, and Mexico have been less inclined to appeal antidumping deci- sions to the courts for a variety of reasons. practices; it contains no recommendations. report. However, responsible officials were consulted during the review, and their views were incorporated where appropriate. GAO/TUSIMWlw) lntenutoxul Wade contents Executive Summary Chapter 1 8 Introduction How Governments Deal With Bumping Antidumping SystemsVary 8 10 DissensionLimits Progressin GATT Antidumping 11 Negotiations Objectives,Scope,and Methodology 12 Chapter 2 15 Sufficiency of ConcernsFocuson Decisionsto Investigate Sufficiency of Evidence Provisions Open to Interpretation 15 16 Evidencefor Initiating U.S., EC, and Mexican Screeningof Petitions Criticized 18 Investigations ConcernsAbout Determining “Standing” 22 Conclusions 23 Chapter 3 25 Improving Antidumping Transparency Requirements Protection of Confidential Information Complicates 25 29 Transparency of the Transparency Goals Antidumping Process tinclwions 33 Chapter 4 35 Appeal Rights Are Administrative Reviews Vary by Country Additional Level of Review in Canada and Australia 35 38 Limited But Still Judicial Review Could Be More Effective 39 Evolving Conclusions 44 Appendixes Appendix I: Antidumping Laws and Practices of South 46 Korea, Brazil, Japan, and India Appendix II: Major Contributors to This Report 48 Tables Table 2.1: Differences in Initial Screeningof Antidumping 17 Petitions and Supporting Evidence Table 2.2: Antidumping Petitions FiIed and Investigations 18 Initiated Table 4.1: Number of Antidumping Casesand Reviews 36 Conducted, 1986-1989 Page 6 GAO/TWLUH1w) lntenuttond Trade Contents Abbreviations CIT Court of International Trade M: European Community GAO General Accounting Office GATT General Agreement on Tariffs and Trade In: International Trade Commission Chapter 1 Introduction Dumping is the sale of products for export at a price less than that charged for the sameor a like product in the “home” market of the exporter. Such international price differences, or discrimination, may occur as a result of businessstrategies that exporters use, including: (1) increasing an overseasmarket share; (2) temporarily distributing products in overseasmarkets to offset slack demand in the home market; (3) lowering unit costs by exploiting large-scaleproduction or learning by doing; and (4) maintaining stable prices during periods of exchangerate fluctuations. These businessstrategies could lead to dumping, as defined by international trade agreements. Views on the harm causedby dumping differ. Economistsgenerally view dumping as harmful only when it involves “predation,” that is, intent by the dumping party to eliminate competition and gain monopoly power in a market. In practice, such predatory dumping has rarely been documented. International trade rules, which take political as well as economicconcernsinto account, view dumping and its potential harm more broadly. These rules define dumping as an unfair trade practice when it “materially”L injures a competing industry in the importing country. The rules provide for the imposition of antidumping duties, or fees, to neutralize the injurious effect of unfair pricing practices. The problem of dumping is not a recent phenomenon.Early in the How Governments 20th century, the United States and Canadaenacted laws to deal with Deal With Dumping this practice. During and after World War I, the U.S. Congressadopted several antidumping statutes. The first U.S. antidumping legislation, the tidumping Duty Act of 1916, required that “predatory intent” be /8” shown. However, soon after its enactment, the act was considered insuf- ficient to protect U.S. producers from dumped imports becauseof the predatory intent requirement. To supplement the 1916 act, the Congress enacted the Antidumping Duty Act of 1921, which forms the basis of current U.S. antidumping laws. This act provided for the application of antidumping duties to offset a margin of dumping.2 The General Agreement on Tariffs and Trade (GAG), which came into effect in 1948, defines the responsibilities and operating rules agreed upon by contracting governments to guide their conduct of international ‘The term “material” la subject to interp~. The Tariff Act of 1930, as amended, defines mate rial lqjury a8 “baml that la not Nconm?quenual, -, or unimpoltant: 2Thedumping~isthepercentbywNchthepricechagedforthesameoralikeproductinthe homemarketoftheexpo&rexc#xlsthe@iportprk!e. P8ge8 GAO/NSlAD&lM) lnt.ernationaI Trade chapter 1 In~ctlon trade relations. The agreement includes a special provision on dumping; the provision does not prohibit dumping, but, rather, allows for a per- mitted responseto dumping in certain circumstances.Article VI of the agreement allows GATT contracting parties to use antidumping duties to offset the margin of dumping, provided that dumped imports can be demonstrated to causeor threaten to cause“material injury” to com- peting domestic firms. Over time, someGATT membersbegan to view other countries’ use of antidumping laws as creating a new barrier to trade. Therefore, during the Kennedy Round of GATT negotiations (1962-1967), the GATT con- tracting parties negotiated an Antidumping Codethat provided a series of rules elaborating on the procedures and methodologiesto be used in applying antidumping duties. Negotiators took up the antidumping issue again in the Tokyo Round of GATT negotiations (1973-1979), partly to provide symmetry with a new agreement dealing with subsidies.The Tokyo Round produced six major agreements,or “codes,” designedto reduce nontariff barriers to trade, including a new Antidumping Code.GATT membersare not required to accept these codesand, in fact, not all have chosento do so. As of Sep- tember 1990, there were 24 signatories to the 1979 GATT Antidumping Code,representing only 35 of the 97 GATT member countries.3 The present code requires determinations of dumping and material injury, as well as a demonstration that dumped imports are causing the injury (“causation”).4 The code provides rules that (1) define dumping and injury, (2) describe the procedures by which signatory governments verify dumping allegations and apply antidumping measures,and (3) provide for dispute settlement. The code also requires signatories to submit semiannual reports of any dumping actions taken during the pre- ceding half year. Our July 1990 report on the use of the 1979 Antidumping Code showed that between 1980 and 1989, Australia, the United States, Canada,and the European Community (EC)initiated 96 percent of the 1,456 new %Ggmtmb of the 1979 Antidumping code are Australia, Austria, Brazil, Canada, Czechoslovakia, the Ewopean Community (lMgium, Denmark, FWKX, Germany, Greece, Ireland, Italy, Luxembourg, theNethertands,portugal,f%mhandthe UnitedKh@onO,&~~C Flnland,How! Kong,Hungw, India, Japan, South Korea, Mexico, New Zealand, Norway, Pakistan, Poland, Romania, Singapore, Sweden, SW, the united states, and Yugaalavia. ‘The Antidumping Code dots not require that dumping be shown to be the primary cause of ~Nury. P8ge 9 GAO/IWAD-9l-59 Intem~tional Trade casesreported.5In addition, Mexico has been relatively active in initi- ating antidumping casessince it joined the code in 198’7,while South Korea and Brazil have initiated considerably fewer cases. The United States, Canada,the EC,and Australia have the most devel- Antidumping Systems oped antidumping systems and practices, while other signatories lack vary practical experience with applying antidumping measures.A number of foreign government officials and experienced attorneys in countries actively using antidumping laws, fearing possible abuseof these laws, have expressedconcern over their adoption and use by newly industri- alizing countries such as Mexico, South Korea, and Brazil. Others, how- ever, believe such fears are exaggerated;they note that the antidumping laws should be seenas an improvement over the systems of onerous trade barriers that they replaced. Information on antidumping procedures in countries other than the United States, Canada,the EX,and Australia is limited. Of the 11 code signatories6designated for our review, Z-Hong Kong and Singapore- have no national antidumping laws or systems. Due to their free-trade orientation, these countries do not support the use of antidumping laws and have not initiated any actions. South Korea, Brazil, Japan, and India have all developed national statutes to deal with dumping. However, Japan and India have reported no caseinitiations to the GATT, and South Korea and Brazil have reported only five and two initiations, respec- tively, between 1980 and 1989. As a result, these countries’ actual antidumping systems and procedures are not well known. Therefore, our analysis focuseson the traditional users of antidumping laws (the United States, the EC,Canada, and Australia) and the new user with the largest number of casesinitiated, Mexico. Appendix I contains a limited discussionof the laws and practices of South Korea, Brazil, Japan, and India. The United States maintains a bifurcated system, established in 1954, whereby separate agenciesdetermine whether the imports are being dumped and whether the dumped imports causeinjury. Currently, the U.S. Department of Commerceis responsible for making preliminary and final determinations of dumping, while the U.S. International Trade Commission(ITC)makes preliminary and final injury determinations. %mathal Track Use of the GATT Antidmphg Code (GAO/NSW90-23SF’S, July 251990). ‘%e United States (for comparison), the EC, Canada, Mexico, Japan, south Korea, Hong Kong, Brazil, singrpore,,md- Pye 10 GAO/MIAJS@l6s Intemtiond Tr8de The Canadian system is similarly bifurcated; however, Revenue Canada, which makes the preliminary and final dumping determinations, is also responsible for making the preliminary injury determination. The Cana- dian International Trade Tribunal only makes the final injury determi- nation. Oneother major difference is that the Tribunal plays an adjudicatory role, whereas the ITC’Sefforts are not only adjudicatory but also investigative in nature. The EChasjurisdiction over the use of antidumping measuresin all member states. The ECchargesone agency,the European Commission, with responsibility for conducting both dumping and injury investiga- tions and making recommendationsto the ECCouncil of Ministers, which has sole authority to impose and collect antidumping duties. ECmember states also play a role in making antidumping decisions,through partici- pation in the Advisory Committee, which advises the Commissionat various stagesduring the proceedings. Although the Australian CustomsService handled all phasesof decision- making in past cases,far-reaching 1988 changesto Australia’s legisla- tion established the Antidumping Authority. The Authority now makes final recommendationsconcerning both dumping and injury (the Cus- toms Service continues to make preliminary determinations of dumping and iqjury). The Authority is an administrative fact-finding body only; it lacks the adjudicatory powers of a tribunal. Finally, Mexico’s antidumping laws, enacted in 1986, give its Secretariat of Commerceand Industrial Development responsibility for making both dumping and injury determinations. Antidumping becamean issue late in the Uruguay Round of GATT negoti- DissensionL&nits ations, which began in 1986. In 1989, countries targeted by antidumping Progressin GATT measures,led by Hong Kong, raised the issue of renegotiating the 1979 Antidumping Antidumping Code during the Uruguay Round of GAIT negotiations. These countries sought to limit the use of antidumping measuresand Negotiations wanted clearer definitions of what constitutes dumping and injury. Signatories started to present proposals in late 1989, and negotiations began early in 1990. Countries agreedthat the vaguenessof the code’s terminology has led to different interpretations of the code and to the development of separate, divergent systems and practices by the major signatories. Proposals have been geared toward harmonizing and imposing more discipline on antidumping practices. P8ge 11 GAO/NSL#B14@ Intemationd Trade However, progress in the negotiations has been extremely limited, and negotiators remain divided on many issues.As of August 1990, parties were unable to reach agreementon a text to be used as the basis for negotiation. One GAIT insider has expressedpessimismover prospects for successfulconclusion of the negotiations. Others have stressedthat the negotiations are extremely complicated and difficult. Although many parties agree on the need for procedural reforms to provide increasedtransparency (open procedures) and adequatejudicial review, a few procedural issues,such as determining a complainant’s standing to bring a case,remain divisive. The most important and contentious issuesin the negotiations involve not procedural but substantive matters: determination of dumping, injury, and causation. Representativesof somecountries targeted in antidumping investigations believe that although improved procedures are important, the fundamental issue is methodology: the rules for determining dumping and injury. If these rules remain unfair, then improved procedures have little value, one official stressed. U.S. industry views of the GATT antidumping negotiations mainly focus on issuesof dumping methodology and are quite polarized. Traditional users of antidumping measures,such as steel producers and semicon- ductor manufacturers, together with the U.S. Chamber of Commerce, opposeany changesthat will weaken current antidumping rules. In con- trast, U.S. exporters and importers, multinational corporations, and computer manufacturers support revisions of antidumping rules, partic- ularly the methods for calculating dumping, to reflect actual pricing practices. In spite of the lobbying efforts of groups seeking changesto the antidumping laws, U.S. officials support the position of the tradi- tional users of antidumping law in the GMT negotiations, and they have strongly resisted proposed changesto the dumping calculation methods. Senators Daniel Patrick Moynihan and Rob Packwood requested that we Objectives,Scope,and compare certain antidumping proceduresof the 1979 GATT Antidumping Methodology Codesignatories who are msjor U.S. trading partners with those of the United States. The objective of our review was to assessthe antidumping procedures and practices of the United States, the EC,Canada,Australia, Mexico, chapter 1 Introduction and, to a lesserextent, South Korea, Brazil, Japan, India, Hong Kong, and Singapore,’ in the following areas: q sufficiency of evidenceto initiate investigations and to make prelimi- nary and final determinations, . transparency to the parties involved and to the public, and l available rights of appeal. To determine the legal requirements for applying antidumping mea- sures, we reviewed the 1979 GATT Antidumping Code and discussedlegal requirements with responsible government officials in the United States and its major trading partners. We also reviewed the antidumping legis- lation of these countries. To identify countries’ antidumping procedures, we obtained and reviewed relevant regulations and procedural guide- lines, when available. We also reviewed law journal articles and recent textbooks discussingthe legislation and practices of these countries. To assessactual antidumping practices, we interviewed government officials from agenciesresponsible for administering antidumping laws, other knowledgeable officials involved with making antidumping policy decisions,and numerous lawyers, consultants, and industry representa- tives who have had experience with antidumping investigations in the United States, the EC,Canada,Mexico, and AustraIia.8 The industry rep- resentatives we contacted included both those who have filed antidumping petitions and those who have been named in petitions. We also met with representatives from Hong Kong, Singapore,Brazil, and the GATT Secretariat to obtain affected exporters’ and GATT officials’ per- spectives on antidumping practices of code signatories and to receive a status report on the GATT Uruguay Round negotiations concerning the Antidumping Code. Our information on antidumping practices is primarily basedon research of available studies and discussionswith knowledgeable practi- tioners. We did not perform any case-specificanalysis of antidumping procedures. As requested, we did not obtain written agency commentson a draft of this report. However, we discussedthe information in this report with ‘Appendix I contains a Iimited discussion of the laws and practicea of South Korea, Brazil. Japan, and India Hong Kong and SIn@pore have IW antidumping kgWatIon and are not discussed further. ‘%rou&out the rest of this report, we refer to these experts, who are Involved with the application of antidumping laws, as antidumping “practitioners.” Page 13 GAO/NSIADB1-33 Intemationrl Trade Department of Commerceand other agency officials during the course of our review and have incorporated their comments where appropriate. Our work was performed from July 1990 through September 1990 in accordancewith generally acceptedgovernment auditing standards. Page14 GAO/IMAD-BlM) IntematIonaI lhde Sufficiency of Evidence for Initiating Investigations The 1979 Antidumping Code and various national regulations allow the five signatories we reviewed a great deal of discretion in determining whether preliminary evidence is sufficient to initiate investigations. This discretion has permitted signatories to establish differing methods for reviewing antidumping petitions and supporting evidenceto deter- mine whether formal proceedingsare necessary.While industry repre- sentatives and trade specialists expressedspecific concernsthat the level of evidence required to initiate an investigation is low in the United States, the EC,and Mexico, fewer complaints were directed at Canada and Australia. We also found that the rate at which antidumping petitions are acceptedfor investigation varies by signatory. Although the code doesnot explicitly require that a petitioner have “standing” in order for an investigation to be initiated (i.e., determining whether petitioning parties represent a major proportion of their industry), practitioners expressedconcern about the perceived laxness in how a signatory determines standing. Determining a complainant’s standing to bring a caseis viewed as one of several important issues being discussedduring the Uruguay Round; however, it remains divisive. key decisionswhich rely, in varying degrees,on the extent of evidence Decisionsto available. These decisions are basedon answering the following Investigate questions: . Should an antidumping investigation be initiated? l Is an affiiative preliminary determination of dumping and resultant il@lry warranted? l Is an affirmative final determination of dumping and resultant injury warranted? Sufficiency of evidence seemsto be the most controversial factor in making the first decision-whether to initiate an antidumping investiga- tion. For preliminary and final determinations, methodology issues(e.g., perceived inequities in calculating the dumping margin) rather than evi- dent&y issuesseemto represent a stronger area of concern. The decision to open a caseis critical becauseof the potentially disrup tive effects antidumping actions can have on trade and becauseof the heavy burden investigations place on respondentsand other parties. Page 15 GAO/NSUD-#lM) Intemationd Trade Antidumping investigations commonly involve requests that foreign exporters and domestic importers fill out detailed questionnaires, which can be as long as 200 pages.Questionnairesmust be completed, trans- lated as necessary,and returned to the investigating authority within about 36 days. Such requests can be particularly burdensome for small companiesthat lack sophisticated bookkeeping records or the resources to engagethe servicesof attorneys or consultants to represent their interests. Parties that elect not to complete questionnaires can be put at a disad- vantage with respect to caseoutcomes.The code states that when par- ties do not provide requested information within a reasonableperiod, authorities may basefindings on the best information available, which may simply be information supplied by the petitioner and the ir@ued industry in support of the dumping allegation. In a recent study, one practitioner noted that the M=,when faced with an uncooperative party, baaedits findings on the least favorable information available, stating that there should be “no bonus for non-cooperation.” When subject to anantidumping action, somesmall businessesmay simply “surrender” to the system by ceasingexports of the allegedly dumped product to the country initiating the action and pursuing other markets where trading policies are more open. A representative of the GMT Secretariat acknowledged that an antidumping investigation involves very comprehensiveand burdensomerequests for information. He further suggested,however, that dealing with such an antidumping investigation can be viewed as a potential cost of participating in another country’s market. Sufficiency of written request that includes evidenceof (1) dumping, (2) injury, and EvidenceProvisions (3) a causal link between the dumped imports and the alleged injury. Opento Interpretation Rather than specifying what information is needed,the code simply says that there must be “sufficient evidence” to support the allegations. However, proposals that attempt to better define the concept of suffi- cient evidencehave been consideredduring the Uruguay Round. In the meantime, the legislative languageused by someof the signatories we reviewed essentially leavesthe issue of deciding what evidence is neededto support an antidumping petition up to the appropriate author- ities. Although the provisions for each signatory addressthe need to substantiate the dumping, ir@ry, and causation elements of an P8ge 16 chapter 3 f3dehcyof-for lnttI8tIngInvamtlgedona antidumping action, what is neededto establish sufficient evidencefor these elements varies by country. Also, somesignatories’ legislation incorporates the provision that the antidumping complaint need only include evidencethat is “reasonably available” to the petitioner. Government officials pointed out that there is no quantifiable, objective standard of how much evidence is “reason- ably available” as well as how much is “sufficient.” Rather, the amount of evidence neededdiffers from caseto caseand dependsupon the nature of the product involved as well as on a myriad of other related factors. These factors include the size of the petitioning company and the accessibility of meaningful product and industry data. We found that the discretion allowed by the code has permitted the sig- natories to use varying practices to review submitted antidumping com- plaints and decide whether to initiate formal investigations. Somekey difference are noted in table 2.1. Tmblo2.1:Diffo~rinlnitkl~olAnt#ufflpkrg P@tiUOlNUfldSupoortkrgEVidOnW EE Canada Awtmlia EiiEi%v Mexico Primary focus of review Dumping Dumping, injury, Dumping, injury, Injury Dumping, injury, and cause and cause and cause Minimum dumoina marain 0.5 5 None used 1-2 None used standare (in &%nt) - Precluded from Yes No Yes No No considering information other than that submitted by the petitioner and facts in the oublic domain Days allowed for initial 20 30 55 OpenC 56 review of oetitionsb %tandard used for identifying cases which may be deemed as too small or frivolous to pursue in light of the time and resources needed to process an antiiumping case. The dumping margin is the percent by which the price charged for the same or a like product in the home market of the exporter exceeds the export price. bThis figure represents the number of days between the filing of an antiiumping complaint and the decision to initiate a case. Some signatories also allow time for a review of the complaint to assure that it has been properly documented. For example, Canada allows an additional 21 days. cT)le EC does not have a statutory deadfine. Estimates of the time it takes the Commission to decide whether an investigation is warranted range from 6 weeks to several months. dAlthough regulations indicate 5 days, of5ciis estimated that the average period between the filing of a complaint and a formal initiition of an antidumping duty proceeding is approximately 3 months. Source: Data provided by off&Is from the five signatories. P8ge 17 GAO/IWADB189 Internationd Trade -2 k?hffl&ncy of Jlvidsneefor Inltla~lnveettgatlona The interpretive nature of the Antidumping Code and implementing reg- U.S.,EC, and Mexican ulations has also allowed for variations in the proportion of filed peti- Screeningof Petitions tions for which investigations are undertaken, as shown in table 2.2. Criticized During our review, practitioners expressedconcernsregarding the min- imal thresholds of evidencethe United States applies in its decisionsto initiate formal antidumping investigations. While similar concernswere expressedregarding the ECand Mexico, fewer complaints were directed at Canada and Australia. Tablo 2.2: AntIdumping FotlUonmFiled and IV lnln8t8d E- !iEz c8n8d# A- conwnunltrc Moxko Period Calendar Fiscal Calendar NA March 1990- year 1989 year 1989 year 1989 August 1990 Number of 28 11 83 NA 8 petitions filed Number of investigations initiated 28 10 24 NA 3 Estimated 1W 91b 38 50 38 percent of formal petitions which result in investiaations Note: NA indicates that the information was either not avaitabk or not applicable. ‘A Commerce official estimated that only 33 percent ta 50 percent of potential antiiumping cases infor- mally brought to the agency’s attention are actually filed. bCanadian data include petitions received during its fiscal year-April 1,1989, through March 31,199o. During this time, Revenue Canada also received 41 “enquiries,” which inch&d telephone calls and letters without supporting documents as well as substantive submissions, The 11 cases noted above represent complaints deemed properly documented. Qccording to a Commission official, the EC does not maintain officii records of the total number of antidumping petitions submitted. He estimated, however, that about SOpercent of the petitions received result in the initiation of an investigatktn. Sources: U.S. Department of Commerce, Ftevenue Canada, Australian Customs Service, European Commission, and Mexico’s Secretariat of Commerce and Industrial Devekpment. U.S. Screening Seenas Initial U.S. reviews of antidumping allegations and supporting evidence are viewed as less stringent than those of other signatories because Least Stringent (1) the Department of Commerceis precluded from considering informa- tion from respondents,(2) the U.S. system focusesduring the pre-initia- tion phase on whether dumping occurxd rather than on the elements of injury and causation, and (3) the Department of Commerceadheresto set time frames for making a decision on initiating an investigation, thus limiting the depth of screening. Page 18 z of lcvidence for lntttulng Invea~tionB Information Considered During its initial review of an antidumping complaint, Commerceis pre- During Initial Screening cluded from considering information from respondents.When deciding to initiate an investigation, Commerceessentially judges each petition only on its own merit, without considering information other than that included in the petition, supporting data submitted by the petitioner, and facts within the public domain. Although this limitation in Commerce’sreview of evidence may seemto result in inadequate screening,it appears that Congressintended that initial reviews be kept simple to avoid burdening petitioners. In an opinion,1the Court of International Trade (CIT)has stated that the Con- gresspurposely restricted the type of information Commerceconsiders in order “to alleviate the burden of petitioners in initiating antidumping proceedings.” Someindustry and foreign government representatives believe that the standard of proof required to compel initiation of a pro ceeding in the United States is too low. U.S. System Fosters Commerce’sapproach to screeningcomplaints is to somedegreethe result of the bifurcation of the U.S. system. Under this system, petitions Initiation are filed simultaneously with Commerce,which has primary responsi- bility for assessingthe extent of dumping, and with the ITC,which has primary responsibility for assessinginjury. The statute directs Com- merce to determine within 20 days whether the allegations in the peti- tion have all the elements necessaryto impose a duty and, if so, whether to commencean investigation. Commerceis thus responsible for deciding whether information reason- ably available to the petitioner concerning material iqiury is contained in the petition. However, rrc, rather than Commerce,has jurisdiction over and expertise regarding matters dealing with material injury. Com- merce officials explained that although the decision to initiate an inves- tigation is essentially made by Commerce,Commercewill usually obtain informal assurancefrom ITC(which acts in an advisory capacity during this initial phase of the process)that there is reasonableevidenceof iIljUIy. A common criticism of the bifurcated system in the United States involves the limited consideration of whether there is a causal link ‘3 Court of International Trade 110,538 F. Supp. 418 (lBSZ), aftied in part and reviewed in part 1 CA IT 39,706 F.2d 1663 (1983). Page 19 GAO/NSUMlM) Intemationd Trade chapter 2 Saffldency of Mdence for Inttiadng Investigations between the dumped imports and the alleged injury. One foreign govern- ment official expressedconcern that since dumping and injury are han- dled by two separate agenciesand, in effect, assessedin isolation, causality tends to be assumedrather than subjected to scrutiny. A for- eign trade expert expressedsimilar concerns.He suggestedthat dumping and injury matters be viewed together. Otherwise, it is less likely that the dual decisionmakerswill have a sufficient overview and appreciation of the injury context within which to make a fully informed and objective decision, particularly on the key issue of causation. The relatively short time frame allowed for making decisionsto initiate formal investigations in the United States, and Commerce’sstrict adher- enceto this deadline, is also thought to hamper the degreeof screening that may be given to submitted petitions. A Commerceofficial stated that somescreeningof potential petitions takes place even before petitions are filed. He estimated that for every two or three casesthat may be brought to Commerce’sattention, only one is ultimately filed. Oncea caseis formally filed, however, a vast majority are acceptedby Commercefor investigation, according to the official. Another Commerceofficial further noted that, in general, the United States may not go into as much depth as other countries in its review of evidence for initiating an investigation. ITCsimilarly finds very few, if any, of the petitions to be insufficient at the initiation stage.According to statistics from Commerce,an estimated 144 antidumping petitions were received from 1986 through 1989, and only 6 caseswere dismissed during this 4year period. EC Screening fU30 Viewed The ECis also consideredless stringent than someof the other signato ries in its initiation of antidumping investigations, although for different as Less Stringent reasonsthan those cited for the United States. The EXis viewed as focusing almost exclusively on the extent of evidenceof injury during the initial screeningprocess,as requiring a low regulatory threshold of evidence for substantiating dumping, and as having an administering authority that exercisesa great deal of discretion in deciding whether to initiate formal proceedings. In the EC,the purpose of screening antidumping petitions before initi- ating an investigation is to ensure that the complaint is (1) admissible, Page 20 GAO/YWIAD@l-b8Intemational Trade Chapter2 Sufficiency of Evidence for lnitLdIlg Inventigations (2) made on behalf of the EC industry affected, and (3) complete, pro- viding sufficient evidenceto justify the initiation of a full proceeding. Initial EC screeningof petitions focusesprimarily on determining injury and, according to industry representatives, the level of data required to establish injury is extensive. A government official explained that more rigorous standards are applied in this area since complainants can be expected to supply full details of the injury suffered by the industry, especially with the assistanceof trade and/or industry associations. Trade association representatives, in turn, estimated that it may take about 6 to 9 months to develop the proposal to initiate an antidumping investigation; data for up to 4 years are requested to be compiled. On the other hand, the EC’Sthreshold regarding evidenceof dumping tends to be low. Government officials explained that they do not place as much emphasis during the pre-initiation phase on determining dumping since petitioners usually do not have accessto information neededto fully establish and document the extent of dumping; this information includes, for example, knowledge of a competitor’s pricing policy in a foreign country or of his production costs. The EClooks more closely at the calculations of and verification for dumping during the investigation phase. During the initial screening,the ECalso tends to be less strict about the need for providing evidence of causality between dumping and injury. Often, simply demonstrating that dumping and injury occur simultaneously is sufficient. Unlike the Department of Commerce,the ECCommissionis not prohib- ited from obtaining and reviewing information from affected parties other than the petitioner during the initial review of the filed complaint. Onepractitioner noted that the ECrepresentatives will sometimessolicit information from the affected domestic industry. Another practitioner believed that the EICCommissionis at least willing during the pre-initia- tion phase to listen to representations of caserespondentsand is not as concernedwith preventing the appearanceof conducting improper “backroom deals” as is the United States. The Commissionis believed to exercise a great deal of discretion regarding various aspectsof the antidumping process.For example, one member state had recommendedagainst holding formal proceedingsfor about six caseswhich the Commissionstill opened for investigation. All were subsequently closed without measuresbeing taken. Page 21 GAO/NSIADB1-52International Trade Foreign government representatives raised concernsregarding the lim- ited extent of review and scrutiny that the ECgives to antidumplng peti- tions and suggestedthat ill-founded allegations can readily pass such a cursory review. In one case,for example, the M: initiated an antidumping investigation against a country which did not produce or even have the capability to produce the raw material in question. In another case,an antidumping investigation was initiated basedon a petition which contained multiple mathematical errors in the com- plainant’s calculation of the dumping margin. Mexican Initiation of Mexico initiated its first antidumping casein 1987 and since then has initiated a total of about 30 cases.In general, Mexico’s antidumping reg- Formal Investigations ulations are viewed by somepractitioners as being vague and open to Leads to Concern use as a protectionist device. U.S. industry representatives described Mexico’s corresponding antidumping practices (including those directed at deciding whether or not an antidumping investigation is warranted) as loose and inconsistent. Factors that may have contributed to this critical view of Mexico’s ear- lier antidumping practices include the newnessof the system, the lack of implementing guidelines regarding the review of antidumping cases,and the Mexican administration’s earlier trade philosophy of protecting its infant industries. A Mexican government official stated the current administration has introduced more “toughness” into its review of dumping allegations and the level of evidenceprovided to support the unfair trading practices. Accordingly, the number of antidumplng cases initiated in Mexico has declined. During 1989, Mexico initiated only 6 antidumping cases,in contrast to 17 casesinitiated in 1987. The Antidumping Code doesnot explicitly require “standing” to file a ConcernsAbout complaint and initiate an investigation. Instead, the code stipulates that Detelmining requests to initiate antidumping investigations be made by or “on behalf “Standing”, of the industry affected.” During our review, however, practitioners expressedconcernsregarding the general laxness with which signato- ries determine whether the complainant represents a sufficient propor- tion of the industry and thus has “standing” to file a complaint and initiate an investigation. They contend that casesin which the petitioner does not represent a major proportion of its industry can be argued to contain insufficient evidenceof injury. Page 22 GAo/NsLm-#182 rnt.ematioMl Trade chpter 2 Saffldency of Evidence for lnMdllglnvesdgatt0n.s A GAIT official referred to standing as one of several “substantive” issuesof discussion during the Uruguay Round. Oneproposal being con- sidered, for example, attempts to define the term a “major proportion” of the industry to represent a specified percentageof the total domestic production of the product comparable to the dumped item. Although Mexico is the only signatory we reviewed that has established a formal minimum threshold, or percentage,of industry production in order to determine standing, this threshold has been criticized as being too low. Mexico’s laws essentially require that the petitioners be respon- sible for producing at least 25 percent of the domestic production of the product comparable to the one allegedly dumped. The United States, on the other hand, assumesthat standing exists for any petition filed unless a majority of the industry shows opposition. A U.S. Commerceofficial explained that when the remainder of the industry (aside from the direct petitioner) has knowledge of a petition and does not opposeit, the petition is consideredsatisfactory and in compliance with the requirements of GATT. Petitioners usually provide information to show what proportion of the industry they represent, and Commercewill generally accept the data presented.Also, Commerce has not established any formal or informal standards for what consti- tutes a major proportion of the industry. What constitutes a “major proportion” of the industry has also not been precisely defined in the EC,Canada,or Australia. An ECCommissionoffi- cial estimated that, in practice, the minimum threshold in the ECwas about 40 percent. Canadian officials recalled that they have initiated several antidumping caseswhich involved even less than 40 percent of the industry. In contrast to the U.S.’ approach of assuming standing, we were told that both the M: and Canada actively verify (by, for example, obtaining input from industry specialists and using available trade sta- tistics) that the majority of the domestic industry supports the petition. Conclusions involve the sufficiency of the threshold of evidenceused to determine whether to initiate an investigation and the laxness in practices for determining whether petitioning parties represent a major proportion of their industry. Proposals being consideredduring the Uruguay Round that attempt to better define the conceptsof sufficient evidence and Page 22 GAOAUtUAD4lds Intenutiond Trade N-2 sumchcyofEvidureefor lnltletinglIlvestlgatlona standing will not alleviate these concernsbut will help provide more gui- danceto new users of the Antidumping Code as well as to the traditional USWS. P8ge 24 GAO/N6IAD414W Intemationd lhde --- Chapter 3 Improving Transparmcy of the Antidumpbq Process The 1979 Antidumping Code requires that signatories adopt procedures that are equitable and open, or transparent, regarding the processof investigating antidumping cases.Despite someimprovement, procedures used to make determinations are still secretive, according to informed sourcesin most of the signatories. Thesesourcesalso express concerns about the authorities’ reluctance to fully disclosethe basesfor making antidumping decisions.Concernsabout the secretivenessof the process are heightened by the lack of accessinterested parties have to confiden- tial information used to develop antidumping cases,especially casesini- tiated by the EC,Australia, and Mexico. These countries do not allow even limited disclosure of confidential information, as do the United States and Canada.The United States has made a number of proposals in the Uruguay Round of GATT negotiations intended to increasethe transparency of the processby improving the quality, amount, and time- liness of information made available to parties during an antidumping investigation. When authorities decide to initiate an ,mvestigation, the Antidumping Antidumping Code requires that a public notice be published and that notice also be Transparency given to various interested parties, incIuding the firms subject to the Requirements investigation, the petitioner, and any other affected importers or exporters. The code also requires that opportunities be given to the peti- tioner, importers, exporters, and the governments of exporting countries to seeaII nonconfidential information that is relevant to the petition and to make presentations to the authorities on the basis of this information. In addition, the Antidumping Coderequires that all foreign suppliers and other interested parties be aIlowed to present evidence and have opportunities to confront and rebut parties with adverse interests during antidumping investigations. Also, public notice is to be given of any preliminary or final finding. In the caseof affirmative findings, such notices are expected to set forth the findings and conclusions reached on all issuesof fact and law consideredmaterial by the investi- gating authorities, and the reasonsand basesfor the decision. Implementation of U.S. procedures are widely believed to be more transparent than those of the other four signatories. This transparency is a function of several Procedures Varies factors, including the U.S.’ public hearings, comprehensiveand publicly available antidumping decisions,and adequate disclosuresof the bases of decisions.Australia and Canada also have fairly transparent proCedUreS* Pa#e 22 GAO/AMA.D@169Inter~tional Trade Conversely, antidumping caseparticipants criticized the proceduresof the ECand Mexico for their lack of transparency, although they noted that procedures have improved in the EC.Much criticism was directed toward the lack of transparency of Mexico’s antidumping procedures, although to somedegreethis lack may be related to the fact that Mexico only recently becamea signatory to the Antidumping Code;therefore it has not had an opportunity to fully develop the necessarymechanisms for a transparent system. In their attempts to comply with the requirements of the Antidumping Code,the signatories we examined have tried to set up various mecha- nisms to make their antidumping procedures more transparent. The mechanismshave not always been uniformly implemented, however, resulting in varying degreesof transparency. The mechanismsinclude l notifying parties to an investigation, . disclosing antidumping determinations, . exchanging information between parties, and l holding meetings of parties to an investigation. Notification of Parties In conformance with the Antidumplng Code,the signatories we examined have established policies to notify parties about when an investigation will be undertaken. In addition to being notified, the par- ties (i.e., affected importers, exporters, and domestic producers) are asked, usually by meansof a questionnaire, to supply information related to the investigation and are given the opportunity to provide submissionsin their defense.The government of the affected exporters is also notified. Four of the five signatories reviewed require that at least a preliminary finding of dumping be established before any antidumping duties are levied. In casessubject to procedures administered by these four signa- tories, respondent firms do have an opportunity to defend themselves before any antidumping duties are levied against them. Mexico’s laws do not reflect this approach, however, since its regula- tions allow provisional duties to be levied within 6 working days after a petition is acceptedand an investigation is initiated. Although on sev- eral occasionsMexico has levied duties before exporters have been given a chanceto defend themselves,Mexican officials have stated that they have stopped this practice. P8ge 26 GAO/NSlADB1M) Inter~cionrl Trade wr3 ImprovhgTruupuencyofthe Antidumping Procem Disclosure of Antidumping Roth Canadian and U.S. investigating authorities provide details about Determinations their calculations and determinations in disclosure meetings with involved parties. Antidumping caseparticipants appear to be generally satisfied with the information on determinations provided by U.S. authorities and the Canadian International Trade Tribunal. In contrast, ECprocedures for disclosing determinations were cited numerous times for their lack of transparency, despite recent improve- ments. Under procedures that the ECadopted in 1988, the Commission publishes relatively comprehensivestatements of the underlying rea- soning for its decisions in the Official Journal of the European Commu- nities. The Commissionalso disclosesfurther information regarding the rationale for its decisions upon request. However, to obtain further information, interested parties must submit written requests and must specify the particular issueson which information is sought. According to two attorneys involved with ECantidumping cases,while the under- lying rationale for determinations is much better known now than in the past, concern still exists that not enough information is provided. This continuing concern may be related to the fact that accessto certain information regarding the basis of a decision, such as the methodology used, is sometimes limited. Criticism was also directed toward the lack of transparency in Mexico’s procedures for disclosing the basesof determinations. Mexican regula- tions provide that parties affected by an antidumping complaint are entitled to information made available to the government. A Mexican government official noted that the administering authority must also show that it has the proper elements to justify making a positive deter- mination of dumping, including elements evidencing dumping, injury, and a causal link. Mexican authorities, however, have not always contacted parties in a timely fashion or provided an adequate explanation of their decisions. Moreover, according to a Mexican government official, authorities will only discusstheir methodology in general terms. Internal documents and computations are not shared with parties. A US. proposal in the Uruguay Round seek to ensure that signatories routinely disclosethe rationale for their antidumping decisions.Onepro- posed change would require that published decisionsset forth the facts and conclusionsof law on which a decision is basedin sufficient detail so that the reasoning is clear. Another proposed changewould require, in the caseof a preliminary or final determination of dumping, that the Pyre 27 GA0/NsIAD@15@ lntern8tiond Trule chtpttr3 b1prodnglban6puenc~ of the .4ntldlunplngRoewr include public hearings, meetings between parties to an investigation, and individual meetings between one party and the investigating authority. Mexico is the only signatory that doesnot provide for meet- ings between parties. However, at the time of our review, Mexico report- edly was in the processof implementing public hearings. In the ECand Australia, meetings of parties, which provide opportuni- ties for parties to state their caseand make rebuttal arguments, can be requested. These meetings are closed in the ECbut open to the public in Australia. In both signatories, however, such meetings have rarely been held. According to an attorney in the EC,such “confrontational” meet- ings tend to be counterproductive, since no real dialogue occurs and the respondent is generally forced to assumea defensive stance. During an investigation, meetings can also be held between the antidumping authorities and individual parties in somecountries. The purpose of these meetings is to give parties the opportunity to provide evidence on their behalf. For example, in the ECat least one such meeting takes place during the courseof an investigation. The meeting is conducted in an informal atmosphere, and no official records are kept. In contrast to the EC,the United States routinely prepares official records of meetings. Both the CommerceDepartment and the ITCare required to put in the public record information such as the identity and affiliation of all persons present and a summary of the factual informa- tion submitted. The Antidumping Code recognizesthat someinformation supplied by Protection of parties to an investigation may, by its nature, be considered confiden- Confidential tial; therefore the code provides that such information will not be dis- Information closedwithout prior authorization. What constitutes confidential information, however, is vague and open to interpretation. Many com- Complicates plaints about the lack of transparency in signatories’ procedures revolve Transparency Goals around interested parties’ inability to seeconfidential information in their adversaries’ submissionsand in antidumping decisions.This lack of accessmakes it difficult for parties to adequately understand and rebut the facts of the original complaint and the rationale for the resulting decision. Furthermore, while parties may be required to pro- vide nonconfidential versions of confidential submissions,antidumping caseparticipants have expressedconcernsabout the usefulnessof these “nonconfidential summaries.” ptgt29 0A0/lwAD-91ds lntmn8tiond Trtdt E=zl Tmuputncy AntldamplneProtut of the Complaints about the lack of procedural transparency appear to be fewer in the United States and Canada,where steps have beentaken to permit legal representatives of involved parties to share confidential information. The U.S.’ negotiating position before the Uruguay Round includes proposals that signatories recognizein the code the desirability of establishing administrative protective order procedures and accept certain amendmentsto the codethat would lead to providing more useful nonconfidential summaries. Protection of Confidential A comprehensivedefinition of what constitutes “confidential informa- Information Obscures tion” is not included in the code. It only states that an example of confi- dential information is information that, if it were disclosed,would be of Basis for Antidumping significant competitive advantage to a competitor or would have a sig- Decisions nificantly adverse effect upon the supplier or the source of the informa- tion. Agreements not to divulge such information have a negative effect on the level of transparency in antidumping determinations and result in a lack of adequate disclosure of the baseson which decisionsare made. Antidumping caseparticipants told us that the transparency of the antidumping decisionmaking processeshas improved in most of the sig- natories reviewed during the past few years. However, with the excep tion of the United States and the Tribunal in Canada,complaints continue to be made regarding the availability of sufficient information for parties to adequately understand the basesof antidumping decisions. For example, an attorney told us that while the transparency of the decisionmaking processhas recently increased in Australia, published determinations continue to be inadequate to understand the basesof a decision. These determinations are basedon confidential information, such as pricing data and calculations, which is not divulged. A respon- dent to a current antidumping proceeding in Australia also expressed similar concernsabout the adequacy of information. A Canadian attorney told us that nonconfidential versions of antidumping determinations carried out by RevenueCanada are useless. Another antidumping caseparticipant stated that while the authority feels that it must provide better nonconfidential information in deci- sions, in actual practice disclosure of information underlying determina- tions continues to be inadequate. Pytao GAo/7muD-s18BxnterMti4BMlTNde chtpttr 3 lmprovlng Trmuptrtncy of the Antidumping Pmcaa Use of Nonconfidential According to the code, parties providing confidential information may Summaries be requested to submit nonconfidential summaries of this information to the appropriate authorities. These nonconfidential summaries are then made available for adversarial review. Antidumping caseparticipants, however, regularly criticized nonconfidential summaries as a poor sub- stitute for supplying the underlying information. With few exceptions, antidumping caseparticipants in Canada,Australia, and the ECcharac- terized these summaries as marginally useful at best and, at times, worthless. Although Mexico has a legislative requirement for supplying nonconfi- dential summaries, a Mexican government official told us that in prac- tice nonconfidential summaries are not prepared, The absenceof such summaries tends to further reduce the level of transparency in Mexico. The Use of Administrative Roth the United States and Canada appear to have made some progress Protective Orders to in reconciling the need for transparent procedures with the need to ade- quately protect confidential information. The system in the United Protect Confidential States, which usesadministrative protective orders, allows confidential Information information to be disclosedto an attorney or other representative of a party to the case.These disclosures are backed up by pledgesof confi- dentiality and by the threat of imposing sanctions in caseof violations. However, somepractical difficulties, including questions of cost and administrative workability, may make it difficult for other countries to institute a similar system. Antidumping caseparticipants and government officials in all signato- ries we examined expressedfewer complaints regarding the trans- parency of the U.S. and Canadian systems. There have been complaints, however, about RevenueCanada, which, unlike its sister agency,the Canadian International Trade Tribunal, doesnot routinely provide accessto confidential information. The United States adopted the use of administrative protective orders following passageof the Trade Agreements Act of 1979. Before 1980, when confidential information was first made available to adversaries’ legal representatives, proceedingswere secret. This secrecy led to what one published source characterized as “proceedings where parties were frequently shooting in the dark.” In 1988, accessto confidential infor- mation was broadened with respect to information submitted to the ITC. P8gt 31 GAO/7vsuDeldB lnttm8tion8I lhdt Attorneys representing both complainants and respondentsin U.S. antidumping casesconsider the investigation processto be relatively transparent. According to one attorney, accessto information in the antidumping processis key for participants, and, in his opinion, the . * admnustrative protective order system is the “single largest advance” for getting fair results. According to another attorney, the system in the United States works well: Any problems with the system usually stem from the difficulty of controlling such large quantities of information. Complaints concerning the Canadian antidumplng system’s lack of transparency largely involve the activities of RevenueCanada,which is responsible for making dumping determinations. Although Revenue Canada is legally authorized to releaseconfidential information to involved parties even if nonconfidential summaries prove inadequate, it rarely does.In contrast, the Canadian International Trade Tribunal, the agency responsible for the injury determination, provides accessto con- fidential information through a system similar to the one in the United States. Attorneys active in the Canadian market give the Tribunal high marks for the transparency of its procedures. Other Signatories’ While developing an administrative protective order system in the EC, Mexico, and Australia may be one way to increasethe transparency of Difficulties in those signatories’ overall antidumping procedures, it is not clear that Implementing an their respective processeswill readily incorporate such a system. Prac- Administrative Protective tical difficulties may stand in the way. Order System One difficulty concernsthe inability of the ECto provide effective sanc- tions against parties in caseof unauthorized disclosure of confidential information. The importance of providing effective sanctions is high- lighted in the United States, where the CommerceDepartment generally will aIlow disclosure only to attorneys, who are subject to disbarment from practice in the event of a violation. According to European govem- ment officials, in the EC12 national laws and 12 separate legal systems exist, with no centralized infrastructure to effectively enforce and mon- itor the preservation of confidentiality of information. According to an ECrepresentative, a seconddifficulty in establishing the use of a&nin&rative protective orders is that individual ECfirms tend to be more secretive and, therefore, are uncomfortable about sharing such sensitive information with outsiders. The firms are concernedthat an administrative protective order system may be abused,may lead to Pt#t8a GAO/N~lM) rnterxutiorul Tr8dt cbtpttr3 Imprwlne-P-~of~ Antidumping Pntttw unauthorized disclosure of confidential information, and may become legally burdensome. As for Mexico, given its current level of development and economic problems, it may have difficulty justifying and implementing what is consideredto be an expensive system. Mexican officials further added that the concept of allowing administrative protective orders is foreign. They believe it would take considerablechangeand effort to have such a concept incorporated into its system. Attorneys in the United States agree that the administrative protective order system has increasedthe transparency of the U.S.’ overall antidumping process,but at a very high cost for all parties involved. In fact, even in the United States one government official was not convinced that the benefits outweighed the costs. Although an Australian attorney noted that Australia has no apparent constraints against implementing an administrative protective order system, Australian officials told us that they have not considered adopting one. U.S. Proposals in the The United States is proposing several amendmentsto the Antidumping Code to strengthen the obligation on investigating authorities to provide Uruguay Round parties with adequate accessto information. The amendmentspropose Concerning that (1) a person claiming confidential treatment for information be Nonconfidential required to show “good cause” for such treatment, not merely “cause”; Summaries (2) the obligation on investigating authorities to provide a nonconfiden- tial summary be mandatory, not discretionary; and (3) the code specify that the summary have sufficient detail so as to permit a reasonable understanding of the substanceof the confidential information. The pro- posal does not state that the code should require countries to adopt such systems in order to increasethe transparency of their antidumping pro- cedures.It does, however, recommendthat a footnote to the code be amendedto recognizethe desirability of establishing administrative pro- tective order systems. Although the transparency of antidumping procedures followed by the Conclusions maor signatories has increasedin recent years, practitioners say the antidumping systems of most of the signatories still lack sufficient pro- cedural openness.The greatest amount of criticism is directed toward the failure of signatories to adequately discloseinformation related to Pyt33 GAO/NSIAIb9149 Inttmtrional Trade industry submissionsand the rationale for making antidumping determi- nations. To somedegree,the authorities are constrained from adopting procedures that are completely open and transparent becauseof the need to protect confidential information. Despite this constraint, there appears to be room to increasethe amount and quality of information made available to parties to an investigation without compromising con- fidentiality requirements. U.S. proposals in the Uruguay Round of negotiations addressmany of the concernsover procedural opennessand due processidentified in this chapter. If accepted,the proposed changesshould help to increasethe transparency of the antidumplng procedures followed by the five major signatories. Ptgt34 GAo/NslAD-e169 lnttrn8tionrl Trdt Chapter 4 Appeal Rights Are Limited But Still Evolving Although the Antidumping Code does not require signatories to provide appeal rights to affected parties, the antidumping systems of the signa- tories we examined allow for both administrative and judicial review. Administrative review in this report means(1) examining whether antidumping determinations should be modified or revoked on the basis of changedprices or circumstancesand (2) considering questions about specific importations. We found that administrative review systems in the countries we examined vary in importance, frequency, and practice. Judicial review, which involves appeals of antidumping issuesbefore courts, also varies in prevalence and importance. Although U.S.judicial review is consideredto be more extensive than that of other countries, somesignatories are considering changesto their systems that will pos- sibly expand the scopeof their judicial review. The United States, the EC,Australia, Canada, and Mexico all provide for Administrative administrative reviews of both dumping and injury determinations. ReviewsVary by Administrative reviews are typically carried out by the agency that con- country ducted the original determination. Proceduresfor initiating and exe- cuting these reviews vary by country. For example, while all the signatories we examined undertake reviews on request by interested parties, Canada and Australia also provide for automatic reviews of dumping determinations by the authorities.1 Differing Importance of The importance of administrative reviews of antidumping determina- tions also varies in different countries. Administrative reviews are par- Reviews ticularly important in the United States for two reasons:(1) They are neededto determine the actual amount of duty owed2and (2) the U.S. system lacks a “sunset provision,” which would end antidumping mea- sures after a fixed amount of time. Although the existence of sunset pro- visions may lessenthe significance of administrative review in Canada and the EC,reviews of specific importations can be important because duties are determined on a case-by-casebasis in both countries. Admin- istrative reviews are probably least important in Australia, where antidumping measuresexpire, with no possibility of extension, after 3 years. lRtvtnut Ctnada auttmaticaUy xwcabha the value on which the duty is based approximately once a year, while the Au&r&an Customs 8ervice reeulprly worms “midttnn” reviews of dumping cxlmbhs18month8aitera llk?amthasbetnimpased. ~tn$eawhdcpaacsbaaedonthedutyratea estimaWintheinitialinv~ons,andthe CommercedireasUS.~to~the~tofdutyoverpaidorasaessany addkhaldutksdue,m deamhedinthe~erevlew. p*= GAO/-NSUDSlU@ lnttrnatlond ‘lhdt Chapter 4 Apped lU@ts Are Limited But Still Ekolv@ Frequency of As shown in table 4.1, the United States and Canada carry out a large Administrative Reviews number of administrative reviews, while the ECand Australia undertake far fewer reviews. This disparity is likely the result of a number of fac- tors, including the importance of reviews in the United States, Canada’s automatic review of antidumping duties, and the existence of sunset provisions in the JX and Australia. Tablo 4.1: Numbor of Antidumping Cases and Revvkwr Conducted, 1M-1 @89 Now cams RWieWS TOM United States 141 355 4% Canada 76 255 331 European Communitya 87 42 129 Australia 120 8 128 Mexicob 30 0 30 aNumbers may be understated because reports do not include cases against nonsignatories. bMexico’s report for the last half of 1969 was not available. Source: Signatories’ semiannual reports to the GATT Committee on Antidumping Practices, 1966-1969. In the United States, requests for Commercereviews of dumping deter- minations are fairly common.According to a recent study, for example, reviews were requested for approximately half of the antidumping orders eligible for review in 1987.3However, practitioners complain that reviews of dumping determinations are not completed in a timely fashion and that the resulting backlogs are a problem since reviews are neededto establish actual duty amounts. A Commerceofficial acknowl- edgedthat delays are a significant problem. However, he noted that part of Commerce’sbacklog was inherited when responsibility for handling dumping determinations was transferred from the Treasury Department to Commercein 1980. Foreign government representatives also have voiced concern that Com- merce administrative reviews place a renewed burden on the respon- dents to provide more information. Practitioners have expressedmore concern about the information requirements of the United States than about those of other countries. Canada’s administrative review processis similar to that of the United States, with a few exceptions. For example, in contrast to Commerce’s practice, RevenueCanada automatically recalculates the value on which the duty is based approximately once a year. 31ncontrast, requests for International Trade Ckmmission administrative reviews of injury are not a3 common, and revocations based on changed circumstances are even less common. PIgt 36 GAO/N8IAD91-69 Intematiod Trade The EChas a smaller number of administrative reviews than the United States for several reasons:(1) The EC’Sregulations contain a sunset pro- vision; (2) adjustments to duty rates resulting from reviews are appli- cable only to future, not past, duties; (3) the ECCommissionenjoys a margin of discretion in conducting reviews, in deciding whether to accept an application for review, and in determining whether to use information provided by parties, according to ECpractitioners; and (4) the EXCommission takes a long time to conduct reviews.4Becauseof these practices, affected parties in the ECdo not often seek administra- tive review. Effects of Sunset Countries that incorporate sunset provisions in their antidumping laws receive fewer requests for administrative reviews, since antidumping Provisions measuresautomatically expire after a fixed amount of time. The routine expiration of antidumping measuresappears to eliminate the need for many reviews. Australia’s sunset provision is particularly strict. It not only eliminates antidumping measuresafter 3 years, but it also provides no possibility for extension. If Australian firms want continued protection against dumped goods, they must reapply for a new inquiry. Sunset clausesin Canada and the ECboth provide for the expiration of antidumping measures6 years after their imposition but also allow for “sunset reviews” on request. In such instances,if interested parties submit information indicating the continued need for duties, measures can be maintained. Sunset reviews are requested for only a small pro- portion of the measures-that expire, according to Canadian and EC sources. The lack of a U.S. sunset provision has concernedsomeforeign repre sentatives. Antidumping attorneys have suggestedthat the United States should adopt a sunset provision (1) to eliminate the problem of backlogs in Commerceadministrative reviews and (2) to allocate the costs of respondents and petitioners more equitably. The present system is thought to put a heavy burden on res,pondents,who must bear the expenseof annual Commercereviews. 4m &W pnnzeding may beBP timeaMsuming aathe initialinv~n, which is normally not les8than6months. Page 87 GAO/lWAD-Bl4l@ lntematlond Trade However, an attorney who represents U.S. petitioners argues that the high cost of initiating an antidumping casein the United States would make establishing a sunset provision unfair to petitioners. Although this argument might be supportable when antidumping measuresautomati- cally expire, as is the casein Australia, the argument seemsweaker if there is a provision for a sunset review, which allows for the extension of measures,if justified. The experienceof Canada and the ECdemon- strates that under the sunset provision, a majority of measuresare allowed to lapse. Several signatories have proposed in the GATT Uruguay Round negotia- tions that a sunset provision be added to the Antidumping Code.Specifi- cally, it has been recommendedthat antidumping measuresexpire after 6 years, with a provision for reviews and extensions. Roth the Canadian and Australian antidumping systems provide for Additional Level of another level of review of decisionsmade by RevenueCanada and the &View in Cmada and At,m-di~ CUSTOMS Servile.These reviews are not ~anied out by the Australia agency that made the initial determination or by the courts. Under the two countries’ procedures, appeals of decisionsto forgo an investigation and appeals of negative preliminary determinations are made to the Canadian International Trade Tribunal and the Australian Antidumping Authority, respectively.6Their decisions are binding on the agency that made the initial decision. The Tribunal and the Authority function somewhat differently. The Tri- bunal is asked only to review whether or not the evidencebefore Rev- enue Canada disclosesa reasonableindication of material injury. The Authority’s role is somewhat broader: It ascertains whether the Austra- lian CustomaService arrived at a lawful finding basedon the material provided for the inquiry. The Tribunal and Authority may only consider the information used by the agenciesto make their decision; other infor- mation or arguments may not be presented for consideration.6 %I Canada, review by the Tribunal can occur at another phase of the investigation: After Revenue Canada haa decided to initiate an inv-, the Deputy Mhiater, exporter, importer, or govem- ment of the country of export may refer the question of a reasonable indication of material injury to theT!ibunal. %ome xepmmtativea of responder& in Canada who are unable to provide information countering the abgatbm of i@ury prior to the de&ion to initiate an inve&@tion have expressed concern over thialimitstioh GAO/NSLiD-@lM) htemtional We chapter4 Appeal Rig&a Am Limited But Still Evolving Parties have not sought Tribunal or Authority review very often. In Canada, there were four such referrals to the Tribunal in 1988, six in 1987, and four in 1986. In Australia, the Authority received three requests to review decisionsnot to initiate investigations during 1988, its first year in operation. The antidumping systems of the United States, the EC, Canada,Aus- Judicial Review Could tralia, and Mexico all provide for someform of review by a higher-level, E3eMore Effective “federal” court. However, judicial review appears to be rather limited in most countries. It is limited because(1) the scopeof review is restricted, thus reducing the number of appeals to the courts; (2) the courts tend to defer to the decisionsof administrative agencies;and (3) someparties (in the EC)are directly prohibited from making appeals. Judicial review by an independent entity is consideredimportant becauseit provides an additional meansof checking the actions and lim- iting the discretion of administering authorities regarding antidumping investigations. Judicial review may be particularly significant in coun- tries where decisionmaking in antidumping procedures is less transparent. Experts believe that U.S.judicial review is more effective than that of other countries becauseU.S. courts seemto have interpreted broadly both their jurisdiction and their scopeof review. In addition, parties are more likely to bring appeals of antidumping to the courts for review in the United States than elsewhere.Other systems of judicial review con- tinue to evolve, and somecountries are considering changesthat could allow for more effective review by the courts. Although judicial review of antidumping decisions has becomemore -widespreadin the major user countries, it is a relatively new develop- ment. In the United States, for example, a judicial review processfor antidumping caseswas not well established until 1980.7 In the current Uruguay Round negotiations, both the ECand the United States have made proposals dealing with judicial review, but these have been rather broad. The U.S. proposal, for example, calls on signatories to “maintain judicial, arbitral or administrative tribunals or procedures for the purpose...of the prompt review and correction of administrative actions relating to final findings...and reviews.” ‘See Judicial Review of Antidumping and Countervailing De&ions, (GAO/NSIAD-84-29, June 1984). Page 39 GAO/NS~169 International Trade Chapter 4 Appeal Bighta Are Limited But Still Evolving In most countries we examined, courts have interpreted their scopeof review to be primarily limited to issuesof law and not factual determi- nations. The US. Court of International Trade is an exception. U.S. Court Better Able to The U.S. Court of International Trade is unique becauseit specializesin Consider Cases handling international trade cases.Therefore, the CIT is better equipped to becomemore involved in the details and facts of an antidumping case. The CIT’s ability to remand a caseback to the administering authority to reassessspecific points enablesit to be more involved with the factual details of an investigation. The ability of a court to consider factual determinations in a caseis important in antidumping investigations, since controversies largely focus on factual issues.U.S. practitioners believe that, as a result of its expertise and its approach, the CIT is less deferential to administering authorities than other appellate courts and more likely to challenge their decisions.However, Commerceofficials also noted that the CIT infrequently overturns an agency’s decisions completely. The CIT has been criticized for unevennessin its deferenceto adminis- tering agencies.Another apparent shortcoming of judicial review by the CITis the length of time the court takes to make decisionson antidumping appeals, according to foreign sources. Although other countries we reviewed currently lack courts that are able to perform detailed review, this situation may change.Many EC lawyers, dissatisfied with the present system of judicial review, support giving the newly created Court of First Instancejurisdiction over antidumping appeals. This court hears appeals in antitrust casesand may be better able to deal with the facts of a case.However, a proposal to give the Court of First Instance competencein antidumping appeals was not acceptedby the EZCouncil, which choseto reconsider the issue after the new court has been in operation for 2 years. The Australian Administrative Review Council has also recommended that the Administrative Appeal Tribunal be empowered to review on their merits all final dumping determinations. According to an Austra- lian Customs Service official, as of August 1990 no decision had been made on the Tribunal’s authority. Pqge 40 GAO/NSIAD91-69 International Trade . Cb8pter 4 Appeal RIghta Are Limited But Still Evolving Limited Scopeof Judicial Parties in Canada,the EC,Australia, and Mexico may bring appeals to the Federal Court of Canada,the European Court of Justice, the Austra- Review lian Federal Court, and the Federal District Court of Mexico, respec- tively. Practitioners told us that the scopeof review by these courts is primarily limited to issuesof law. As a result it has beenobserved that the courts may be reluctant to challenge antidumping decisionsmade by administering authorities. In the ECin particular, the Court’s approach has given the Commissionbroad discretion in making its determinations. Consequently, many ECpractitioners told us, the Court often tends to uphold the Commission’sdecisions. Judicial review by the U.S. Court of Appeals for the Federal Circuit is comparable to judicial review by the courts described above. CIT deci- sions are appealable to the Court of Appeals for the Federal Circuit and ultimately may be appealed to the SupremeCourt. Somepractitioners believe that this appellate court lacks the expertise of the CIT and tends to defer to the decisionsof the administering authorities. In particular, a U.S. attorney stressedthat becausethe factors for determining injury are not so concrete and are subject to interpretation, the appellate court may be reluctant to challenge the rrc’s decisions. Accessibility and Coverage All interested parties are able to requestjudicial review in most coun- tries. However, in the EC,the European Court of Justice has interpreted of Judicial Review the provisions of the Treaty of RomeQso as to prohibit independent importers from appealing casesdirectly to the Court. Instead, these importers are entitled to challenge a specific importation in ECmember state courts, and these courts may then refer the issueto the European Court of Justice. This indirect accessto the European Court of Justice has several limitations: (1) Arguments may becomedistorted, because the casesare filtered through national courts; (2) time limits within which the casemust be referred to the Court are not established; and, most notably, (3) caseswill not always be referred to the Court. In most countries, judicial review is limited mainly to determinations that are considered final, except in the caseof Australia.9 The Austra- lian system allows for appeal of both positive and negative decisionson initiation as well as preliminary and final determinations. *Although IX antidumping laws do not contain provisons for judicial review, the Treaty of Rome (which &abliihed the EC) includes provisions for judicial review. QA party in Mexico, however, recently appealed a prelii determination on constitutional grounds. P8ge 41 GA0/NSIAlM1-58 international Trade Clmpter 4 Appeal Righta Are Limited But StiJl Evolving Judicial Review Parties in Canada, Mexico, the EC,and Australia have been less likely than U.S. parties to bring appeals casesto the Court for several reasons. Prevalence and Practitioners in Canada and the EChave noted that the limited scopeof Importance judicial review keeps them from bringing appeals cases.In Canada,par- ties are also not likely to appeal for other reasons,such as (1) the expenseof bringing appeals, combined with Canada’ssmall market size, which reducesthe perceived benefits of appealing, (2) the lack of con- sultants who understand the law well, and (3) the lack of necessary information to appeal RevenueCanadadecisions,according to lawyers and officials. Similarly, exporters to Mexico may be reluctant to appeal becausethey sell only small amounts in the Mexican market. In addition, experienced practitioners considerjudicial review in the EC to be ineffective becausethe Court (1) has infrequently overturned the Commission and (2) has taken an extremely long time-often 2, and as many as 4, years-to make decisions. Although the slow pace of reviews and the small market size may deter Australian parties from bringing appeals, another factor may encourage them to make more appeals. In Australian appeals cases,parties are able to obtain confidential information that was not available to them during the course of the investigation. As a result, someAustralian experts believe that appeals to the Court, which historically have not been common, will increase. It is difficult to assessthe impact that judicial review has had on antidumping practices. In the United States,judicial review has contrib- uted to the interpretation of U.S. antidumping laws and regulations. It appears that CIT’Sscrutiny has had significant effect on Commerceand the ITC. U.S. officials and practitioners claim that judicial review deci- sions affect the way that the agenciesconduct their work. For example, if the courts tell the ITCto consider a particular factor in making its injury determination, it will do so for all subsequentinvestigations, to avoid another challenge in the courts. The impact of judicial review has been less definite in other countries. Somepractitioners note that judicial review has established legal prece- dents and improved procedures. But others, particularly in the EC,are skeptical about the effect of review by the courts, since decisionsso often uphold the determinations of the administering authority. Page 42 GAO/NSLUMl-lW International Trade Chapter 4 Appdlu#ItaheLbuIt4!dButst#lIBvoIvIng Other Possible Avenues of Other opportunities for appeal do not provide a practical alternative for most parties. Appeals to the CATI‘ cannot be brought by private parties, Appeal and the binational panel created by the United States-CanadaFree Trade Agreement deals only with casesinvolving parties from the two countries. Although the Antidumping Code provides for review, signatories have not used this processvery often. In 1988,two cases(one from Japan and the other from Sweden)were referred for conciliation and settlement to the GAIT Antidumping Committee. Somepractitioners believe that use of the GAG dispute settlement mech- anism will increase in the future. An experienced EClawyer believes that Japan may appeal to the GAIT more frequently as a result of the ineffective ECappeals process.Also, an ECofficial noted that provisional changesto GATT procedures have made the dispute settlement process faster. He also observedthat the composition of the dispute settlement panel is likely to work in favor of exporters, since most GAIT members are neither users nor supporters of antidumping measures.1o Nevertheless,others believe that small countries may be reluctant to request review from GATT.One representative from a smaller exporting country voiced concern over possible negative political repercussions that could result from such an action. The United States-CanadaFree Trade Agreement provides for another review option for American and Canadian parties to appeal antidumping casesinvolving the two c~untries.~~The agreementprovides for reviews by a panel consisting of knowledgeable practitioners from the two nations with expertise in the antidumping area. The panel is supposedto apply the samestandard of review and the samelegal principles as domestic courts, and its decisions are binding on both governments. It is difficult to assessthe panel’s impact becauseit has consideredonly a few cases.However, several practitioners said that panel members have tended to interpret the panel’s mandate broadly, suggestingthat the panel will examine antidumping issuesin much greater detail than a typical appellate court would. lOHowever,the parties in a dispute .xttlment cam approvethe compasStlon of the panels. 11TheFreeTradeAlgeementalsorim,to~anen~forbochtheUnitedstateSand Canada,replacing antidumpinglawsw&hcompetftlon leuaNegoUatomhavebeen given 7 years to reachthisgoal. Pyre 43 GAO/NEIAMlbQ Intemational ‘hde Ch8pter 4 Apped Rights Are Limited But Still Evolving Administrative reviews are less important in other signatories than in Conclusions the United States, where they are neededto set the actual amount of duty owed. The United States has been criticized for its lack of a sunset provision in its antidumping law, as well as for its slowness in carrying out annual reviews of dumping determinations. Signatories have pro- posed that the Antidumping Codebe changedto address both these issues. In most countries, judicial review is limited to issuesof law. The U.S. Court of International Trade is an exception becauseits specialized function makes it better able to addressthe details of a case.Although judicial review in other countries may be less thorough, the systemsof the other signatories we examined continue to develop. Recentproposals aimed at more complete judicial review in the ECand Australia indicate that substantial changesare possible. In the GATT Uruguay Round negotiations, both the United States and the EChave made broad proposals introducing judicial review to the Antidumping Code.However, it may be difficult to effect changes through the Antidumping Code,becausejudicial review doesnot fall within the traditional scopeof antidumping rules. P8ge 44 GA0/lvsuDe149 lntem8tlon8l Tr8de l P4e 45 GAO/NSIAD91-69 Intemationd Trade Appendix I Antidumping Laws and Practices 6f South Korea, Brazil, Japan, and India While South Korea, Brazil, Japan, and India have antidumping legisla- tion, few caseshave been initiated and processedby these countries. South Korea has initiated five investigations, none of which has resulted in the imposition of a duty. Brazil has initiated two investigations, both resulting in the application of a duty.* Neither Japan nor India has reported any formal antidumping investigations. India and Japan have, however, received one and three antidumping petitions, respectively. We compared the antidumping legislation of these four countries with the GATT Antidumping Code,specifically with respect to provisions for sufficiency of evidence,transparency, and available rights of appeal. In light of what we believe to be the very broad and interpretive nature of the code,the implementing regulations seem,in general, to conform with the GAIT Code and addressthe basic tenets of issuesunder review. Each country’s regulations, for example, addressthe need to ensure that there is “sufficient evidence” to support the dumping, injury, and causation elements of an alleged antidumping action. Bach also acknowledgesthe need, in varying degrees,to establish transparency in their antidumping systems and to provide for a meansof reviewing decisionsmade by the administering authorities. In the absenceof sufficient caseexperience, it is difficult, if not impos- sible, to determine what procedures and policies these countries would adopt given an increase in the number of antidumping complaints filed. Similarly, the degreeto which a country’s antidumping processesare consistent, transparent, and not subject to manipulation can only be evaluated in the light of actual experience. According to an attorney currently representing a U.S. company involved in a South Korean antidumping action, South Korea’s pre-initi- ation stage appears to be informal. However, parties seemto enjoy con- siderable accessto information during the investigation as well as the pre-initiation phasesof the antidumping process.The attorney did express concern about South Korea’s seemingly low threshold of evi- dencefor initiating an investigation. In his particular case,the South Korean government decided to initiate an investigation despite the fact that imports were decreasing,domestic production of the imported article was rising dramatically, and domestic production was signifi- cantly undercutting the price of imports, he explained. lIntemational Trade:Useof the GATTAntidumping code (GAO/NSIAD-90-238FS, July 25. 1990). Page 46 GAO/NSIAB91-59 Intem~tio~I Trade In Brazil, the Customs Policy Commission has responsibility for making both dumping and injury determinations. Oncean investigation has been initiated, the Commissionsendsout questionnaires to potentially affected parties. However, according to a Brazilian government official, resource constraints make it very difficult to conduct subsequenton-site verification of submitted data. He also noted that Brazil generally pre fers and encouragesresolving casesthrough price undertakings2because of their administrative convenience. The very limited experience under the Japaneseantidumping provisions suggeststhat the pre-initiation stage may be more important in Japan than it is in the United States and may be viewed as an important oppor- tunity for parties to negotiate an agreement.For each of the three antidumping petitions filed in Japan, actions were terminated in the pre- initiation stage after the Japaneseindustry obtained (1) a “voluntary export restraint” agreement from the exporters limiting the amount of the allegedly dumped product sold to Japan or (2) other assurancesby the appropriate exporting industry to take corrective action. Imple- menting guidelines in Japan specifically state that an investigation may not be initiated when measuresto eliminate an alleged injurious effect to an industry in Japan have been taken. Indian legislation provides that before deciding to initiate an investiga- tion, the government must satisfy itself that it has “prima facie” evi- denceof dumping, injury and, where applicable, causality. However, with no casesto serve as a context, this standard gives no indication of whether the threshold for initiating an investigation would be any more or any less strict than that in the United States. 2A price undertaking isbasically a voluntary measure by an exporter to increase prices or to cease exports at dumped prices to elhinate the ir&riow effect of dumping. Page 47 GAo/NsIADa-s9 Intem8tion8l Trde --- _ Appendix II Major Contributors to This Report John Watson,Assistant Director National Security and Jill Derderian,Evaluator International Affairs Division, Washington, DC. Kane Wong, Issue Area Manager SanFrancisco Evelyn Aquino, Evaluator-in-Charge RegionalOffice Robert Tomcho, Evaluator Gerhard Brostrom, Writer-Editor Kimberly Dillon, Intern Richard Perruso, Attorney Office of the General Counsel (4ss661) Page 48 GAo/NsxAD-g159 Intemational Trade
International Trade: Comparison of U.S. and Foreign Antidumping Practices
Published by the Government Accountability Office on 1990-11-07.
Below is a raw (and likely hideous) rendition of the original report. (PDF)