Much has been written about the failure of the U.S. antidumping law to live up to its rationale, even as articulated by the law’s supporters.1 Likewise, many of the methodological quirks and institutional biases inherent in the law’s administration have been documented and described in previous studies.2
Despite antidumping’s egregious flaws, the law garners broad bipartisan support in the U.S. Congress. But only the most cynical of policymakers could continue to support the application of antidumping measures when their intended purpose — to remedy injurious dumping — ceases to exist.
Under the World Trade Organization’s Antidumping Agreement and under U.S. law, a “sunset review” (also known as an “expiry review” or “five‐year review”) to determine whether an antidumping measure is still needed to redress injurious dumping must be initiated within five years of the measure’s original imposition.3 But a rule that was intended to prevent indefinite application of superfluous import restraints has proven ineffective. During seven years of rendering so‐called sunset reviews, the U.S. antidumping authorities have revoked fewer than one in every four measures where a domestic interest favored continuation.
In April, the U.S. International Trade Commission voted by a margin of four to two against revoking antidumping (and countervailing duty) restraints on imports of hot‐rolled steel from Brazil, Japan, and Russia, despite compelling evidence that the domestic industry is thriving according to every conceivable measure of business performance. The facts of that case are accessible enough to illuminate some of the problems with the sunset process. It is apparent that reform of the process will be required to rein in the abuse that interminable antidumping restrictions represent.
Now You See the Sunset…
Before the conclusion of the Uruguay Round of trade negotiations (1986–1994), which produced the World Trade Organization and, among other agreements, the Antidumping Agreement (ADA), some countries with antidumping laws had no formal procedures for terminating restraints imposed pursuant to those laws. There were no requirements that existing measures be reviewed to determine whether they were still warranted.
The problem was particularly acute in the United States. At the end of 1994, 32 U.S. antidumping measures originally imposed before 1980 were still in effect without any intervening assessments of the conditions of the domestic industries involved.4 What were originally justified as temporary measures to counter the injurious effects of dumping morphed into permanent fixtures that distorted trade and subsidized protected producers.
The ADA set out to fix this problem. Article 11.1 of the ADA stipulates: “An anti‐dumping duty shall remain in force only as long as and to the extent necessary to counteract dumping which is causing injury.“5 To ensure that national authorities honor this commitment, Article 11.3 of the ADA further stipulates: “any definitive anti‐dumping duty shall be terminated on a date not later than five years from its imposition … unless the authorities determine … that the expiry of the duty would be likely to lead to continuation or recurrence of dumping and injury.“6
…Now You Don’t
While the sunset provisions of the ADA were hailed as a liberalizing achievement, the fact is that they have done little to ensure termination of superfluous antidumping measures in the United States. The U.S. track record of revocation is especially bleak.
Between July 1998 (the month of the first U.S. sunset review determination) and May 2005, 335 antidumping sunset review determinations were issued. In 80 of those 335 cases, the measures were revoked because there was no domestic interest in their continuation.7
In the remaining 255 cases, where a domestic industry demonstrated interest in preserving the antidumping measures, the Department of Commerce found that revocation would be likely to lead to a continuation or recurrence of dumping in every case. And in 194 of those 255 cases, the ITC found that revocation would be likely to lead to a continuation or recurrence of material injury.8 Thus, based on the accumulated record to date, the likelihood of revocation pursuant to the U.S. sunset provisions is nonexistent at the DOC and only 23 percent at the ITC. These statistics seem to contravene the intended reform negotiated during the Uruguay Round.
To put those numbers in perspective, consider the revocation statistics in other large, antidumping jurisdictions. Over a comparable time frame — cases initiated between 1998 and 2004 — the Canadian International Trade Tribunal has revoked 40 of 85 antidumping orders pursuant to expiry reviews, which is a revocation rate of 47 percent, more than twice as high as that in the United States. 9 Likewise, during the same period the European Union revoked 36 of 82 cases subject to expiry reviews, a revocation rate of 44 percent.10
Although these comparative data show that U.S. antidumping authorities are much less likely to revoke an antidumping measure than their Canadian and European counterparts, they do not help to explain why that is the case. A brief review of the laws, regulations, and other administrative procedures surrounding U.S. sunset reviews might help with that question.
Shell Games and Fortune Tellers
It does not require a great deal of investigative prowess to comprehend why the DOC always finds that revocation is likely to lead to a continuation or recurrence of dumping. According to its established procedures for conducting sunset reviews, the DOC normally will determine that dumping is likely to continue or recur if
dumping continued at any level above de minimis [0.5 percent] after the issuance of the order or the suspension agreement, as applicable;
imports of the subject merchandise ceased after issuance of the order or the suspension agreement, as applicable; or
dumping was eliminated after the issuance of the order or the suspension agreement, as applicable, and import volumes for the subject merchandise declined significantly.11
Meanwhile the conditions that would normally lead to a determination of no likelihood of continuation or recurrence of dumping are limited to the case where “dumping was eliminated after the issuance of the order or the suspension agreement, as applicable, and import volumes remained steady or increased … In analyzing whether import volumes remained steady or increased, the Department normally will consider companies’ relative market share.“12
Thus, the only conditions that normally would warrant a “no likelihood” finding are conditions that are extremely unlikely given rational economic behavior. The foreign exporter would have to increase his U.S. price — so that it is no longer dumping according to the DOC’s slanted methodologies — and then find customers willing to purchase the same or greater volume than it sold previously at lower prices. This chain of events plainly violates established properties of a normal demand curve.
Many of the conditions necessary for a “no likelihood” finding are simply beyond the control of the foreign exporter. Given the prospective nature of U.S. antidumping duty liability assessment,13 importers are the entities that assume the risk when purchasing from a foreign producer that is subject to an antidumping order. Importers can never be sure of their ultimate financial obligations to the U.S. Customs Service. Thus, the decisions of risk‐averse importers to forego purchases from subject foreign producers are ultimately what determine those producers’ sales volumes in the United States.
In many cases, foreign producers have a difficult time convincing importers to continue purchasing precisely because of the uncertainty presented by the antidumping measure. Expecting the importer to discount the risk and continue purchasing, but at a higher price, is akin to a jobless person with no assets or credit history asking for a prime rate loan. Success at either is highly dubious.
Furthermore, conditions that normally lead to a “likelihood” determination can be false positives. The cessation or decline of imports after the issuance of an order can reflect many factors, such as difficulty finding U.S. importers willing to continue purchasing, the relative attractiveness of other markets, or prohibitively high antidumping rates.
Accordingly, the bright line test specified in the policy bulletin is necessarily distorted by decisions beyond the control of the producer subject to antidumping restraints and cannot be considered a useful barometer of what might happen if the restraints were revoked. It is a veritable shell game.
Ultimately though, a sunset review can end in revocation regardless of the DOC’s determination. In fact, 23 percent of all contested sunset reviews ended in revocation despite the DOC’s findings.
Is it a reasonable expectation that 77 percent of domestic industries protected by antidumping measures likely would be injured within a foreseeable time if those measures were revoked? The more likely explanation for the paucity of revocations lies in the fact that the ITC is obligated statutorily to gaze into a crystal ball and project what might transpire should the measures be revoked. Prospective assessments are problematic enough — financial experts lose money in the stock market every day — but to mandate that officials speculate, opens the door unnecessarily to subjective interpretations preconceptions, and political influence.
The recent decision in the sunset review of hot‐rolled steel from Brazil, Japan, and Russia provides some rich insights into the problems created by this prospective statutory construct.
The Greatest Hot‐Rolled Steel Case on Earth
The hot‐rolled steel case is instructive because it features a split vote — four to two in favor of continuation — and thus avails differing opinions from the same factual record. It also tells a story of just how difficult it is to get an antidumping order revoked.
The original measures were imposed in 1999, when U.S. steel production was relatively fragmented; many producers were having difficulty remaining viable; bloated labor contracts, pension and healthcare obligations, and other inefficiencies were dragging down profitability; and a flood of imports had been diverted to the strong U.S. market because Asian steel demand had nearly completely collapsed in the wake of its widespread financial crisis in 1997 and 1998.
Today the situation is dramatically different. The U.S. steel industry has undergone structural consolidation, shedding inefficient, uncompetitive mills; unloading most of its legacy costs on the taxpayer‐backed Pension Benefit Guarantee Corporation; negotiating new flexible labor contracts; consolidating operations; and achieving a degree of market power that has enabled it to dictate terms to its customers. The year 2004 was characterized by record industry profits, persistent steel shortages, perpetual price hikes, raw materials surcharges, and occasional broken contracts by steel producers who reckoned they could get higher prices on the spot market by reneging on their commitments. Meanwhile, worldwide demand has experienced a massive outward shift over the past few years, primarily because of the growing appetite for steel in underdeveloped countries like China, India, and those of Eastern Europe. A strong global steel demand for many years to come is almost certain.
The majority and dissenting opinions of the ITC in the sunset review do not diverge over this now‐and‐then characterization of the hot‐rolled steel industry. It is supported by ample evidence on the record and acknowledged throughout the determination. But there is a considerable divergence of opinion with respect to what these data mean. For example, the majority opinion found that the “industry’s greatly improved performance in 2004 … resulted mainly from highly favorable market conditions, including unprecedented increases in global demand, tight global supply, and sharply rising prices.“14 It went on to conclude, “In short, the factors that enabled the domestic industry to achieve strong operating results in 2004 are not likely to continue. Thus, while we do not consider the domestic industry to be currently in a weakened condition, it is susceptible to the continuation or recurrence of material injury.“15
By contrast, the dissenting opinion considers the industry stronger and fundamentally changed. “The depth and breadth of these changes all indicate that the condition of the domestic industry is much changed, and much improved, from the period of the original investigations. The industry’s excellent performance in 2004 further supports this conclusion. It is unlikely, then, that revocation would affect the industry in the same way and to the same extent that subject imports affected the domestic industry during the original investigations.16
The majority opinion reflects the view that 2004 was serendipitous; that the structural changes in U.S. and world supply and demand will have no or little bearing on the long‐term health of the industry; and that if restrictions were lifted, imports from the subject countries, attracted by relatively higher U.S. prices, would flood back into the U.S. market in high volumes, suppressing or depressing prices, and negatively affecting the domestic industry. The working assumption in the majority opinion, an argument put forth by the domestic industry, is that increased Chinese demand, which drove the conditions of the world market in 2004, would be satisfied by Chinese production going forward. As a result, the world’s exporters — including the subject producers — would again look to the United States to absorb their collective output.
Pointing to some excess capacity in the subject countries, the majority concludes that “producers in the subject countries have the ability and incentive to send significant quantities of hot‐rolled steel to the United States.“17 Furthermore, the majority finds that “underselling would likely be significant in the event of revocation/termination given subject producers’ pricing behavior during the original investigations, the importance of price, the substitutability of the products, and the fact that increased volumes for this product would likely be achieved through lower prices.“18 But the dissenting opinion sees these data differently — and quite frankly, more objectively.
“Although China continues to build new steel capacity, demand is projected to increase by 8 percent in 2005, in part because of preparations for the 2008 Olympics and the Expo 2010. The record also indicates that there is increasing price parity between the U.S. prices and other major world markets. In short, the U.S. is not the safe haven for steel that it was during the original investigation.
“Overall, the home market demand and export trends in the cumulated countries have changed considerably since the original investigations. While there are existing inventories and in some cases, available capacity, given the worldwide changes in demand, we cannot conclude that it is more likely than not that imports will increase to such an extent as to cause material injury to the domestic hot‐rolled industry.“19
Exiting the Hall of Mirrors
The facts of the hot‐rolled steel case underscore an inherent shortcoming of the sunset process: expecting objectivity from the DOC and forcing predisposed ITC Commissioners into the rolls of soothsayers make for bad policy. It is difficult enough for the ITC, on a retrospective basis, when a factual record exists, to accurately attribute domestic material injury to imports in the presence of other causal factors. To require the ITC to speculate what those data might be and then speculate the effects of those speculated data on the domestic industry is unworthy of the formality of the process. There can be little doubt that under the current system, antidumping measures that should be revoked are continued routinely because of vastly differing interpretations of the statutory criteria and because of the inherent fallibility of decisions that are based on speculation.
One sensible, even‐handed way around the current morass would be to combine automatic termination of antidumping measures with expedited provisions to reinstate them should conditions of injurious dumping recur within a reasonable time frame — say one year. Under this proposal, all antidumping measures would be terminated after five years without exception or deliberation. However, domestic industries would be allowed to file new petitions, as early as the date of revocation, which would require evidence of material injury or threat of material injury by reason of less than fair value imports, just as is required in original petitions.
For petitions filed within one year of the previous order’s expiration, special provisions designed to expedite the process would prevail. An affirmative preliminary injury finding rendered by the ITC within 45 days in accordance with its normal criteria would suffice to reinstate the dumping order with rates prevailing at the time of revocation. In other words, a DOC investigation, which normally takes six to twelve months to produce a preliminary dumping rate, would be unnecessary for reinstituting the dumping order. Under this proposal, relief could be available to domestic industries found to be injured within 45 days of revocation of the original order. After the duties were reimposed, a traditional DOC investigation would take place and would entail a recalculation of the duty rates. This idea was originally proposed in a Cato paper published in 2002.20
The sunset review process in the United States is broken. As a result, antidumping measures that distort trade and impose real costs on interests throughout the economy are allowed to continue well beyond their legal justification. It is imperative that U.S. policymakers fix this problem. Automatically terminating antidumping measures after five years with provisions for expedited re‐imposition if the legal thresholds are met could go a long way toward improving the system.
2. See Brink Lindsey and Dan Ikenson, “Antidumping 101: The Devilish Details of ‘Unfair Trade’ Law,” Cato Trade Policy Analysis no. 20, November 26, 2002.
3. Actually, there is a difference between the WTO language and the U.S. law on this point. The WTO agreement reads that an antidumping measure should be terminated within five years of imposition unless the authorities find that revocation would likely to lead to a continuation or recurrence of dumping and injury. But the U.S. law mandates that a notice of initiation of an expiry review be published no later than 30 days before the fifth anniversary of the antidumping measure. Accordingly, most U.S. decisions to continue or revoke an order are not rendered until the sixth year, which would seem to be inconsistent with U.S. WTO obligations.
4. Michael O. Moore, “Commerce Department Antidumping Sunset Reviews: A Major Disappointment,” George Washington University Occasional Paper Series, October 7, 2002, p. 1, http://www.gwu.edu/~gwcsg/OPS/moore1.pdf.
6. Ibid. Article 11.3.
7. Data compiled from two sources: Import Injury Investigation Case Statistics (FY 1980–2003), U.S. International Trade Commission, Office of Investigations, November 2004, http://www.usitc.gov/trade_remedy/USITC_Stat_Report-11–04-PUB.pdf, and other sunset review decisions and data available at http://www.usitc.gov/.
9. Figures compiled from expiry review decisions and data available on the website of Canadian International Trade Tribunal at http://www.citt-tcce.gc.ca/dumping/reviews/orders/rrin98_e.asp.
10. Figures compiled from antidumping expiry review data appearing in annual reports of the European Commission to the European Parliament titled: “Anti‐dumping, Anti‐subsidy, Safeguard Statistics Covering the Year [1998, 1999, 2000, 2001, 2002, 2003, 2004],“Annex F (for 2001 through 2004 data) and Annex G (for 1998 through 2000 data). Reports are available on the website of the European Union, http://europa.eu.int/comm/trade/issues/respectrules/anti_dumping/stats.htm.
11. Department of Commerce, International Trade Administration, “Policies Regarding the Conduct of Five‐year (“Sunset”) Reviews of Antidumping and Countervailing Duty Orders; Policy Bulletin,” Federal Register, April 16, 1998, p. 18873. The “Policy Bulletin” has been the subject of WTO dispute settlement in a few cases. Most recently, the Appellate Body upheld a dispute panel’s finding that the bulletin is a “measure” subject to WTO dispute settlement, but reversed the panel’s finding that Section II.A.3 of the bulletin is inconsistent with Article 11.3 of the ADA due to what is considered “insufficient analysis” by the panel. (See World Trade Organization, Appellate Body Report, United States‐Sunset Reviews of Anti‐ Dumping Measures on Oil Country Tubular Goods from Argentina, November 29, 2004.)
13. Under the U.S. system, antidumping duties are estimates of duty liability paid by the importer of record to U.S. customs authorities. Those estimates are based on calculations of dumping from entries during a previous period. Only after the merchandise has cleared customs, normally as part of an annual “administrative review,” will the actual liability be calculated. So, if the antidumping deposit rate was 10 percent at the time of entry, but the actual liability was calculated at 20 percent, the importer would receive a bill (likely months or even years after the entry) for the difference between 20 and 10 percent, plus interest. Thus, there is a lot of uncertainty (i.e., business risk) in continuing to do business with foreign exporters that are subject to antidumping measures.
14. U.S. International Trade Commission, Certain Hot‐Rolled Flat‐Rolled Carbon‐Quality Steel Products from Brazil, Japan, and Russia: Investigations Nos. 701-TA-384 and 731-TA-806- 808 (Review), April 2005, p. 28.
15. Ibid., p. 41.
16. Ibid., p. 43.
17. Ibid., p. 36.
18. Ibid., p. 38.
19. Ibid., p. 58 (emphasis added) .
20. See Brink Lindsey and Dan Ikenson, “Reforming the Antidumping Agreement: A Road Map for WTO Negotiations,” Cato Trade Policy Analysis no. 21, December 11, 2002.