The Obama administration has pro-posed changing several Commerce De-partment antidumping procedures. Thosechanges would expand the scope for find-ings of dumping and precipitate a surge inantidumping actions.

The evolution of U.S. antidumping pol-icy is marked by distinct inflection pointscorresponding to legal and administrativechanges favorable to protection-seekers. Inthe decades following World War II, theU.S. antidumping apparatus was graduallycaptured by interested parties, and the lawtransformed from one predicated on pro-tecting consumers and preserving competi-tion into a tool that suppresses competitionin the name of remedying “injury” done todomestic producers.

The arcane mix of statutory rules anddiscretionary whimsy that emerged is a farcry from the first antidumping law—inboth practice and intent. No longer areanti-competitive or predatory pricingpractices the law’s target. Instead, the tar-get is foreign competition writ large, and,as such, antidumping practice routinely

punishes normal, healthy competition atgreat cost to the broader economy.

Honest debate about the law’s purposeand consequences has been stifled by itscomplexity and by the persistence of highlyinaccurate rhetoric about the propriety ofstrong antidumping rules to redress unfairforeign practices. Armed with the pretenseof a noble purpose, representatives of laborunions and import-competing industriesoften cite the number of antidumping initi-ations as evidence of the need for an evenmore accessible and restrictive antidumpinglaw to better protect “us” from “them.” Butthe causation implicit in that logic is back-ward.

This paper shows that the increase inantidumping activity reflects several develop-ments that have nothing to do with foreignbehavior, including a progressive expansionof the definition of dumping, relaxation ofevidentiary standards, and a pro-domesticindustry bias in the law’s administration.Ultimately, the antidumping remedy is amuch larger problem than the dumping it ispresumed to address.

Protection Made to OrderDomestic Industry’s Capture and

Reconfiguration of U.S. Antidumping Policyby Daniel Ikenson

Daniel Ikenson is associate director of the Center for Trade Policy Studies at the CatoInstitute and coauthor of Antidumping Exposed: The Devilish Details of UnfairTrade Law (2003).

Executive Summary

OLI

CY

AN

ALY

SIS

TRA

DE

POLI

CY

AN

ALY

SIS

TRA

DE

POLI

CY

AN

ALY

SIS

TRA

DE

POL

December 21, 2010 No. 44

Introduction

The Obama administration is proposing toamend certain aspects of the CommerceDepartment’s oversight of the U.S. antidump-ing law. Revised methods for estimating majorcomponents of the dumping calculation, high-er thresholds for foreign companies to exceedin order to demonstrate that they are not“dumping,” tighter rules governing methods ofpaying estimated antidumping duties, andother changes have been put forward, ostensi-bly to give the law more teeth.

History suggests that if those proposals areimplemented, a surge in antidumping actions—at great cost to the broader U.S. economy—islikely to follow. The evolution of antidumping ismarked by distinct inflection points correspond-ing to legal and administrative changes imple-mented at the behest of certain import-compet-ing industries and their political allies. As theutility of antidumping became evident to thoseindustries, trade negotiators, and politicians inthe decades following World War II, the law—and its regulatory apparatus—was graduallytransformed from one predicated on protectingconsumers from collusive or predatory foreigntrade practices into a tool to suppress competi-tion in the name of remedying “injury” experi-enced by domestic producers. The arcane mix ofstatutory rules and discretionary whimsy thatemerged as contemporary antidumping policy isa far cry from the first antidumping law—inpractice and intent. Today, antidumping is littlemore than an elaborate excuse for run-of-the-mill protectionism.

No longer are anti-competitive or predato-ry pricing practices the target of the antidump-ing law. Rather, its target is price discrimina-tion—specifically, the act of a foreign firmcharging lower prices in the United States thanit charges in its home market for the sameproduct. But there is nothing inherently wrongor predatory or unfair about such a pricingstrategy. Even if there were, evidence of anti-competitive behavior is never required to initi-ate an antidumping case or to impose restric-tions. Antidumping borrows its rhetoric from

the past, but it presently punishes normal,healthy competition that certain domestic pro-ducers find objectionable.

Honest debate about the law’s purpose andconsequences has been stifled by its complexityand by the persistence of highly inaccuraterhetoric about the propriety of strong antidump-ing rules to redress unfair foreign practices.Armed with the pretense of a noble purpose, representatives of labor unions and import-competing industries often cite the number ofantidumping initiations as evidence of the needfor an even more accessible and restrictive anti-dumping law to better protect “us” from “them.”Granted, U.S. antidumping initiations have in-creased considerably over the decades, but notbecause of an increase in anti-competitive behav-ior among foreign producers.

Rather, antidumping use has increased be-cause the law has been “strengthened” by ex-panding the definition of “unfair” and grantingdomestic industry too much access to the leversof antidumping administration. Meanwhile,collateral damage to innocent victims and thebroader economy continues to mount.

This paper describes the evolution of U.S.antidumping policy from an obscure offshoot ofcompetition law into the predominant instru-ment of contingent protection that it is todayand provides an account of some of the crucialstatutory and administrative changes that haveoccurred over the decades. Its purpose is todemonstrate that the increase in antidumpingactivity reflects several developments that havenothing to do with foreign behavior whatsoever,including a progressive expansion of the defini-tion of dumping, relaxation of evidentiary stan-dards, and a pro-domestic-industry bias in thelaw’s administration at the U.S. Department ofCommerce. Ultimately, the antidumping reme-dy is a much larger problem than the dumpingit is presumed to address.

What Is Dumping?

In economics, dumping is defined as the actof a firm charging lower prices in an exportmarket than in the domestic market for the

2

Antidumping borrows its rhetoricfrom the past, but it

presently punishesnormal, healthy

competition thatcertain domestic

producers findobjectionable.

same product. Dumping is a fancy term forcross-border price discrimination, which itselfcarries a malevolent connotation.

But there is nothing sinister or even unusualabout a firm engaging in price discrimination.Indeed, U.S. firms charge different prices in dif-ferent regional markets for the same goods for avariety of legitimate reasons. For example, anincumbent firm that has operated in NewEngland for many years and whose brand is wellknown and respected by consumers might beable to charge higher prices in that market thanit can charge in California, where it is an un-known entity attempting to enter the market. InCalifornia, the firm might have to induce con-sumers to try its product by offering lower prices.

The same strategy can be employed withoutsinister motive if those markets happen to be indifferent countries. A foreign firm that sells itswidgets in the United States at lower pricesthan it charges at home may be trying to reapthe benefits of the brand it has cultivated andnurtured at home by charging higher prices,while simultaneously pursuing a common market-entry strategy of charging lower priceswhere its brand is relatively unknown. In thatregard, price discrimination is not a problem inneed of a remedy. There are perfectly rational,legitimate, profit-maximizing justifications forengaging in a strategy of price discrimination.Accordingly, dumping is not systemicallypredatory, anti-competitive, or unfair.

However, in some cases when certain condi-tions are present, dumping can be a cause forconcern. A sustained strategy of cross-borderprice discrimination could prove to be anti-com-petitive and welfare-reducing if: (1) the foreignfirm in question has market power (i.e., can setprices in one or both markets); (2) the marketsare sufficiently segregated to foreclose the possi-bility of price arbitrage; and (3) demand for theproduct in the export market is more price elas-tic than demand in the home market. The pres-ence of those conditions could enable the kindof predatory, anti-competitive practices that arecommonly—although mistakenly—presumedto be the target of the U.S. antidumping law.

Without those conditions present, however,a dumping strategy would be unsustainable.

Without sufficient market power to set andmaintain higher prices at home, the firm wouldhave insufficient profits from which to cross-subsidize lower-priced sales abroad. Withouthigh enough tariffs, transportation costs, orother market barriers to sufficiently segregatethe home market, reimportation of the lower-priced imports into the home market wouldarbitrage away any dumping-enabling profits.And if the demand of customers in the exportmarket is less price elastic than the demand ofcustomers in the home market, the price of thegood in the export market could equilibrate ata level higher than the price at home. In otherwords, the home-market profits needed to en-able a dumping strategy might in fact requirehome-market prices that are lower thanexport-market prices, rendering such a strategyself-defeating or impossible.

Although economic theory provides for theseconditions, one would be hard-pressed to findany real-world examples of dumping causingcompetitive markets to become monopolistic.Nevertheless, the U.S. antidumping law is osten-sibly concerned with the kind of internationalprice discrimination that stems from those con-ditions. To be more precise, the antidumping law,according to its defenders, exists to remedy orcountervail the effects of those anti-competitiveconditions, which are presumed to exist becauseof various foreign government policies. A 2002Bush administration submission to the WorldTrade Organization Negotiating Group onRules put it this way:

A government’s industrial policies orkey aspects of the economic systemsupported by government inaction canenable injurious dumping to takeplace. . . . For instance, these policiesmay allow producers to earn highprofits in a home “sanctuary market,”which may in turn allow them to sellabroad at an artificially low price. Suchpractices can result in injury in theimporting country since domesticfirms may not be able to match theartificially low prices from producersin the sanctuary market.1

3

One would behard-pressed tofind any real-worldexamples of dumping causingcompetitive markets to becomemonopolistic.

Thus, antidumping is portrayed as a legiti-mate government response to anti-competitiveforeign government policies. However, the firstmajor shortcoming of that justification is that,under the law, no evidence is required to demon-strate that (nor does the administering authorityinvestigate whether) any of those anti-competi-tive conditions exist—let alone that they are theproduct of some foreign government policies.The existence of price discrimination (i.e., somesales that are priced lower in the United Statesthan in the home market) is simply assumed tobe proof positive of the existence of those anti-competitive conditions, and those conditions areassumed to be the product of foreign govern-ment policies, even though there are plenty oflegitimate reasons for a firm to price discrimi-nate.

Under U.S. law, dumping is defined as thesale of a commodity in the U.S. market at aprice that is less than “normal value.”2 Normalvalue is derived either from the selling price ofthe same or a similar product in a comparisonmarket (normally the home market of the for-eign producer) or from the cost of producingthe product in question, plus allowances forselling expenses, administrative expenses, andprofit.3 The amount of dumping is calculatedby subtracting the export price from the nor-mal value, and the dumping margin, expressedas a percentage, is that difference divided by theexport price. Thus, if a foreign producer sells aproduct for $11 in his home market and for$10 in the U.S. market, the amount of dump-ing is $1 and the dumping margin is 10 per-cent.

For antidumping duties to be imposed, theadministering authorities must also determinethat the domestic industry is materially injuredor threatened with material injury, or that thedevelopment of an industry is materially retard-ed by reason of the dumped imports in ques-tion.

Thus, any price discrimination that causesor threatens injury to a domestic industry is thetrigger for imposing duties—and not just pricediscrimination that arises from anti-competi-tive conditions abroad. Under the law, there issimply no requirement or effort to distinguish

the causes of price discrimination and, as aresult, perfectly unobjectionable business prac-tices are routinely punished at great expense toinnocent victims.

A law that once required close examination ofunderlying market conditions and the motives offoreign sellers now simply grants the worstassumptions without requiring any meaningfulevidence of anti-competitive behavior or theunderlying market distortions that are presumedto give it rise.4

The explosion of antidumping in recentdecades is in no small part attributable tochanges in U.S. law and administration that haveexpanded the definition of dumping to includeroutine and unobjectionable pricing practicesand made antidumping protection increasinglyeasy to obtain. Over the course of about eightdecades, antidumping evolved from an instru-ment of law aimed at preserving competition toa bureaucratic apparatus devoted to restricting it.

How did such an obscure policy instrumentevolve into what some economists consider to bethe most prevalent impediment to internationaltrade today?5 What explains the continuousbroadening of the scope of U.S. antidumpinglaw? Why does a practice that is so antitheticalto free trade have sanction within the rules ofthe World Trade Organization? What explainsthe near total disregard for the interests of con-sumers and consuming industries in the con-temporary administration of the antidumpinglaw? The complex history of U.S. trade politicssheds some light on these questions.

Antidumping’sOriginal Charter

Prior to the signing of the General Agree-ment on Tariffs and Trade in 1947, antidump-ing laws of the United States were premisedupon the logic of antitrust.6 Like antitrust, thepurpose of the antidumping law was to preventunfair competition, which could be the result ofmonopolies engaging in predatory pricing. Forbetter or worse, governments seeking to curtailthe exercise of market power by domestic firmscould turn to their antitrust laws. But it was far

4

Over the course of about

eight decades,antidumping

evolved from aninstrument of law

aimed at preservingcompetition to a

bureaucratic apparatus devoted

to restricting it.

more complicated for governments to apply andenforce their antitrust laws against foreign firms.And since it was apparent that foreign govern-ments might be less thorough in applying theirantitrust rules to their own firms’ behavior in theU.S. market, the U.S. government turned toantidumping.7

Prior to passage of the first U.S. antidumpinglaw in 1916, unfairly traded imports were regu-lated under the Sherman Antitrust Act of 1890,which “declared illegal any effort to combine orconspire to monopolize a particular market,”and the Wilson Tariff Act of 1894, which ex-panded the prohibitions of the Sherman Act tofirms outside the United States. The WilsonAct made it “unlawful for foreign producers tocombine or conspire to monopolize the U.S.market.”8 By federal statute, violations of theselaws mandated criminal prosecution. Anti-competitive practices concerning imports cov-ered under the Sherman and Wilson Acts werepunishable by substantial fines or imprison-ment.9

In 1904, the Canadian government becamethe first to have an antidumping law, when theparliament enacted legislation authorizing theimposition of “a special duty of customs” equalto the difference between the “fair market value”and the “selling price,” whenever “it appears tothe satisfaction of the minister of customs . . .that the export price . . . is less than the fair mar-ket value thereof.”10

The primary purpose of the Canadian lawwas to enable the government to respond tocompeting demands from two vital constituen-cies—agriculture and steel. Agricultural inter-ests wanted a significant cut in tariffs across theboard, while the Canadian steel industry want-ed protection from its principal competitor,U.S. Steel.11 The government worried that rais-ing tariffs on steel would be too risky, as itwould encourage every other interest to lobbyfor higher protection of their industries.

By establishing a “special duty” to be admin-istered separate from the general tariff and at thediscretion of customs authorities instead of thecourts, the Canadian government produced anappealing solution to its dilemma—a mecha-nism to grant contingent protection without

need of revising the general tariff structure. Thatthe first firm investigated for antidumping wasU.S. Steel, a perennial target of U.S. antitrustauthorities, helps explain the lack of indignationon the part of Canada’s trading partners.12 Andthe discovery of antidumping as a vehiclethrough which protection could be doled outconditionally and at the discretion of politiciansproved internationally appealing. Rather thancondemn the Canadian policy, Australia, NewZealand, the United Kingdom, and Franceadopted antidumping policies of their own.13

The United States joined the antidumpingclub in 1916, when the Wilson administrationwas under pressure from industry for an increasein U.S. tariffs on the grounds that Germany wasengaged in economic warfare in the form ofpredatory dumping. Sympathetic to thoseclaims but averse to antagonizing free traders inthe Democratic Party, Wilson asked Congressto put antidumping provisions into the RevenueAct of 1916. Congress obliged, inserting lan-guage that criminalized the importation of mer-chandise at prices lower than “actual marketvalue” when such practices are undertaken with“intent to injure, destroy, or prevent the estab-lishment of an industry in the United States orto restrain competition.”

Like the Sherman and Wilson Acts beforeit, the Antidumping Act of 1916 was a criminalstatute. But unlike the Canadian law of 1904,the first U.S. antidumping law did not grantcustoms the authority to levy special duties ondumped imports. Rather, the U.S. law sought toremedy predatory dumping with severe penal-ties, including punitive fines (triple damages)and imprisonment.14 But demonstrating preda-tory intent under the 1916 Act proved difficult,and few cases were brought.15

In 1919, the U.S. Tariff Commission pub-lished the results of a study on foreign competi-tion, which was informed largely by surveys ofdomestic manufacturing firms. The study rec-ommended that the Antidumping Act of 1916

be revised and strengthened, and thatsome official body, moving along linessanctioned by Congress in the FederalTrade Commission Act, may realisti-

5

The discovery ofantidumping as avehicle throughwhich protectioncould be doled outconditionally and atthe discretion ofpoliticians provedinternationallyappealing.

cally be specifically instructed to dealwith dumping as a manifestation ofunfair foreign competitive methods . . .that in the case of imports, bonds pro-viding for the collection of dumpingduties subsequently assessed may beuseful, and that the President or theSecretary of the Treasury may be em-powered to impose additional duties,or even to refuse entry when industri-ally destructive dumping is proven orimpending.16

Congress adopted the recommendations ofthe Tariff Commission study in the Anti-dumping Act of 1921, which authorized thesecretary of the Treasury to levy “a specialdumping duty” equivalent to the differencebetween the “exporter’s sales price” and the“foreign market value (or, in the absence ofsuch value . . . the cost of production)” when-ever “an industry is likely to be injured, or isprevented from being established, by reason ofthe importation into the United States of for-eign merchandise of such class or kind is beingsold or is likely to be sold in the United Statesor elsewhere at less than its fair value.”17

With passage of the new act, responsibilityfor the enforcement of the antidumping lawwas shifted from the judiciary to the executivebranch, while the object of regulation wastransformed from predatory pricing to injuri-ous price discrimination. Economist J. MichaelFinger characterized the changes to U.S. lawbetween 1890 and 1921: “Trust-busting re-mained the rallying cry, but the object of theregulation shifted from trusts to imports, theinstrument from law to bureaucracy.”18

The impact of the Tariff Commission studywas profound. By elevating injury to U.S. pro-ducers as the primary rationale for having anantidumping law—and in the process marginal-izing concern about the impact of predatorypricing or the imposition of duties on U.S. con-sumers—the report was the catalyst that sparkeda dramatic transformation of the antidumpingregime. The welfare of consumers was no longera concern of the antidumping law. As Fingerputs it: “the ‘injury to competition’ standard

[was] replaced by a ‘diversion of business’ stan-dard, the sort of diversion that is a normal partof the competitive process.”19

Despite the relative ease of obtaining contin-gent protection under the 1921 AntidumpingLaw, U.S. industry was largely shielded fromforeign competition thanks to the high levels oftariff protection afforded by the Fordney-McCumber Act of 1922 and the Smoot-Hawley Tariff Act of 1930. Antidumping wouldbegin to emerge as an attractive alternative afterWorld War II, as the multilateral endeavor toliberalize world trade gained momentum.

Antidumping in the GATT

The U.S. government requested that lan-guage permitting the use of antidumping beincluded in the GATT. In fact, as noted by eco-nomic historian Douglas Irwin, the Anti-dumping Act of 1921 provided the “textualbasis” for the initial GATT rules on antidump-ing.20 Indeed, the WTO Antidumping Agree-ment in effect today authorizes member statesto impose antidumping duties when administer-ing authorities find that imports determined tobe sold at less than normal value are causing orthreatening injury to a domestic industry ormaterially retarding the establishment of anindustry.21

The view widely held—or at least the excusegiven—by U.S. negotiators during the earlydecades of GATT was that the existence ofantidumping made greater trade liberalizationpossible because it assured Congress that certainfallback contingencies were available if generaltariff cutting exposed domestic interests to toomuch competition. Without antidumping, theargument went, Congress might not have sup-ported general tariff liberalization.22 That a dis-tinctly protectionist policy should be consideredan element of trade liberalization was a bit coun-terintuitive. However, that assertion was madeat least partially credible by the fact that the U.S.government exercised considerable restraint inits administration of U.S. antidumping policy.Between 1921 and the end of 1967, over 89 per-cent (631 out of a total of 706) of all U.S.

6

Responsibility forenforcing the

antidumping lawshifted from thejudiciary to the

executive branch.

antidumping investigations ended with negativedeterminations of injurious dumping.23

The U.S. argument that instruments of con-tingent protection could facilitate trade liberal-ization was tested during the Kennedy Roundof GATT negotiations (1964−1967), whenAmerican negotiators found themselves in theawkward position of advocating the expansionof GATT rules to other nontariff trade barriers,while simultaneously resisting the attempts oftheir counterparts to bring antidumping understricter multilateral discipline. Those advocat-ing stronger disciplines on antidumping pre-vailed in the Kennedy Round by agreeing toguidelines for injury determinations that weremore rigid (gave less latitude to the administer-ing authorities) than those applied by the U.S.Tariff Commission, which had taken overresponsibility for injury determinations fromthe Treasury Department in 1954, under U.S.law.24

Congress did not acquiesce to this en-croachment on its jurisdiction, and madeexplicit in the legislation authorizing theKennedy Round agreements that the UnitedStates would be bound only by those provisionsof the antidumping code consistent with exist-ing U.S. law.25 It would not be the last time thatCongress asserted its authority to shield theU.S. antidumping law from reform.

The launch of the Tokyo Round of GATTnegotiations (1973–1979) provided U.S. law-makers occasion to once again revisit—andexpand—the definition of dumping under U.S.law. Until 1974, dumping was defined as thepractice of a firm selling in an export market atprices lower than those it charged in its homemarket. Price discrimination—regardless of itscause or intent—was the target of the U.S.antidumping law. There was no requirement,nor were efforts undertaken, to ascertainwhether the price discrimination in questionhad an innocent explanation. Accordingly, in1974 the deck was already stacked in favor ofdomestic protection-seekers.

Measuring price discrimination usually en-tailed comparing the net prices of the same orsimilar merchandise sold by a given firm in bothmarkets and calculating an average difference, or

averaging net prices of sales of the same or simi-lar merchandise and comparing those averagesbetween markets. Under either approach,though, all home-market sales “made in the ordi-nary course of trade” factored into the determi-nation of the average home-market price. Theidea was to make an apples-to-apples compari-son. But at the behest of various domestic pro-ducer interests, who had been trying for sometime to persuade the Treasury Department of thepropriety of more aggressive dumping margincalculation techniques, Congress amended thelaw to include a provision for a “cost test” in theTrade Reform Act of 1974.26

The purpose of the cost test was to eliminatefrom determination of the average home-market price those home-market sales made atprices below the full cost of production. Onlythose sales made at prices above the cost of pro-duction would now factor into the averagehome-market price, and if less than 10 percentof the sales of a given product “model” weremade at those prices, then all sales of that modelwould be disregarded and the home-marketcomparison would be based on a “constructedvalue” that included approximations for the costof production, selling, general and administra-tive expenses, and an amount for profit. By elim-inating lower priced home-market sales, theeffect of the cost test was to produce higherhome-market average prices or to necessitateuse of constructed values, both of which inflateddumping margins.

The fig leaf of a legal justification offeredfor the change was that sales made at pricesbelow the full cost of production could not beconsidered sales “made in the ordinary courseof trade.” If the ordinary course of trade is for afirm to be profitable, then sales below the fullcost of production must be anathema to thatobjective, as they detract from profit. Whywould a firm sell below cost if its objective is tomake profits?

But basic microeconomic theory holds—andeconomists, cost accountants, and analysts havenoted—that selling at prices below the full costof production is often the profit-maximizingcourse of action for firms, and that selling belowthe full cost of production is clearly a practice

7

The launch of the Tokyo Roundof GATT negotiations provided U.S. lawmakers occasionto once again revisit—andexpand—the definition of dumping underU.S. law.

that is “in the ordinary course of trade.” Manyfirms—particularly those operating in highfixed-cost industries—drop their prices belowthe full costs of production when facing reduceddemand for seasonal or cyclical reasons. As longas the price charged is high enough to cover thefirm’s variable costs, any price above variable costcontributes toward coverage of the firm’s fixedcosts. If the alternative is to stick to a higherprice and sell nothing, then there are no variablecosts to cover, but there is no contribution tofixed costs either. Under those circumstances,the accounting losses would be higher for thefirm.

Consider the production of hot-rolled car-bon steel, which involves the use of iron ore,coke, and other minerals; electricity; water; andother variable inputs—inputs that are consumedin some proportion to output and consumptionof which increases with output. But steel pro-duction also involves high fixed costs—coststhat are incurred regardless of the amount ofoutput—on account of the need for expensiveproduction facilities and the high cost of pur-chasing and operating the necessary capitalequipment. Regardless of whether the factoryproduces one million tons or nothing at all, thetotal fixed costs are the same. All of the fixedcosts have to be incurred in order to produce thefirst ton of steel, which is also the total amountof the fixed cost necessary to produce one mil-lion tons of steel, or more.

If the per-ton total fixed cost is $100 andthe per-ton total variable cost is $100, then ide-ally, the firm would be able to sell at a priceabove $200 and make a profit. But if econom-ic conditions are such that the market will beara price of only $150 per ton, it is still profit-maximizing for this firm to sell at that price,even though it is below the full cost of produc-tion and even though the firm will incur a loss.If the firm does not sell the steel, it incurs a lossof $100 (the fixed cost). If it sells at $150, thefirm incurs a loss of only $50. In this case, profit maximization means loss minimization.Selling below the full cost of production, aslong as the price is above the average variablecost of production, is the rational course ofaction.

Prior to the 1974 act’s mandate of a costtest, the Treasury Department addressed theissue of sales below cost in two different cases.In both, Treasury acknowledged the funda-mental economic principle that selling belowthe full cost of production was sometimes theoptimal decision. Treasury also noted that saleswere made at prices below the full cost of pro-duction in both the U.S. and home markets,which made for a fair comparison.27

Ultimately, domestic industry prevailedupon Congress to adopt a below-cost provisionin the 1974 Act, which mandated that salesbelow the full cost of production be disregardedin the calculation of average home-market prices(but, unsurprisingly, made no such requirementon the U.S. side of the calculation). In caseswhere all home-market sales were at pricesbelow the cost of production (and, thus, evi-dence of a protected, sanctuary home market inwhich high profits could be used to cross-subsidize cheap export sales entirely absent), theforeign firm would be further penalized by bas-ing the dumping calculation on a comparison ofits U.S. prices to “constructed value,” which wasto be the sum of the full cost of production plusthe larger of actual or estimated amounts forselling, general, and administrative expenses,plus a minimum of 8 percent for profit.

The effect of the cost test on the dumpingcalculation can be profound. To demonstrate,in Table 1 there are five hypothetical sales of aparticular type of widget in the U.S. market atprices ranging from $1.00 to $5.00. Likewise,there are five sales of identical products in thehome market also ranging from $1.00 to $5.00.Assuming the volumes sold in each transactionare the same (to simplify the process of weightaveraging the prices), the weighted-averagedprice in both markets is $3.00 and the dump-ing margin should be zero because there is noprice discrimination at all.

But the cost test introduces another step tothe calculation methodology, which produces adifferent end result. The cost test restricts theeligibility of home-market sales that can factorinto the weighted-average price. Specifically,only sales at prices above the full cost of pro-duction factor into the average. Accordingly,

8

Ultimately, domestic industry

prevailed uponCongress to adopt

a below-cost provision in the

Trade Reform Actof 1974.

the two home-market sales priced below theunit cost of production, which is $2.50 in Table1, are eliminated, causing the average home-market price to rise to $4.00. This change gen-erates a dumping margin of $1, or 33 percent,despite the fact that there are no price differ-ences between the markets.

If the cost of production were $5.50 insteadof $2.50, then all sales in both markets wouldbe at prices below the full cost of production.The average U.S. price would still be $3.00, butsince all home-market sales would be ineligiblefor calculation of the average home-marketprice, a “constructed value” would be calculatedto serve as a surrogate for home-market price.The constructed value would be equal to thecost of production plus allowances for selling,general, and administrative expenses (a mini-mum of 10 percent under the statute after the

1974 Act) plus an allowance for profit (a min-imum of 8 percent under the statute after the1974 Act).

At a minimum, the constructed value in thisexample would be $6.49 ($5.50 for cost; $0.55for expenses; $0.44 for profit) and the dumpingmargin would be $3.49 ($6.49 minus $3.00), or116 percent. That is a pretty substantial marginof dumping—particularly for a firm that ischarging identical prices in both markets and isby definition not price discriminating.

A 2002 Cato study exposing the method-ological distortions of the U.S. antidumpingregime included a table of counterfactuals—results that would have been obtained in 18real-life antidumping cases, had particularmethodologies been changed or forgone alto-gether. In 17 of 18 cases, the cost test was em-ployed. Had the cost test not been employed in

9

By mandating acost test under U.S. law, Congressensured that dumping calculations wouldno longer be the product of pure price comparisons.

Table 1

Cost Test (U.S. Dollar)

Net Home-market Net Home-market

Net U.S. Price Price Unit Cost Cost Test Price (used)

1.00 1.00 2.50 Fail --

2.00 2.00 2.50 Fail --

3.00 3.00 2.50 Pass 3.00

4.00 4.00 2.50 Pass 4.00

5.00 5.00 2.50 Pass 5.00

Average 3.00 3.00 2.50 4.00

Table 2

Cost Test (U.S. Dollar)

Net Home- Net

Net U.S. market Unit Cost Home-market Seller’s Constructed

Price Price Cost Test Price (used) Cost Expense Profit Value

1.00 1.00 5.50 Fail — 5.50 0.55 0.44 6.49

2.00 2.00 5.50 Fail — 5.50 0.55 0.44 6.49

3.00 3.00 5.50 Fail — 5.50 0.55 0.44 6.49

4.00 4.00 5.50 Fail — 5.50 0.55 0.44 6.49

5.00 5.00 5.50 Fail — 5.50 0.55 0.44 6.49

Average 3.00 3.00 5.50 — 5.50 0.55 0.44 6.49

those 17 cases, the average calculated dumpingmargin would have been 59.7 percent lower.28

In 5 of the 18 cases, calculation and use of con-structed value was necessary. Had the build-upof constructed value not included an elementfor profit, the margins would have been 11 per-cent lower.29

By mandating a cost test under U.S. law,Congress ensured that dumping calculationswould no longer be the product of pure pricecomparisons—a development that furtherdivorced the legal definition of dumping fromits economic meaning. And, as evidencedabove, it served to inflate dumping marginsand antidumping duty rates, and subsequentlyheightened the appeal of the antidumpingremedy to importing-competing industries.

Regulatory Capture

In addition to the technical changes toantidumping calculation methodology, theTrade Reform Act of 1974 also provided busi-ness a formal channel of access to antidumpingand other trade policymaking. The act man-dated the establishment of a private sectoradvisory committee system to ensure the inclu-sion and incorporation of business viewpointsin the negotiation, monitoring, and implemen-tation of GATT agreements and related poli-cies. The private sector consultation system wasdivided into three vertically integrated compo-nents: the President’s Advisory Committee forTrade Policy and Negotiations; four policy ad-visory committees; and 22 industry-trade advi-sory committees, jointly administered by theDepartment of Commerce and the Office ofthe United States Trade Representative.

As the 1974 Act included language authoriz-ing U.S. participation in the Tokyo Round ofGATT negotiations, the Trade Agreements Actof 1979 implemented U.S. obligations commit-ted during the Round. And it made severalimportant changes to U.S. antidumping admin-istration, including transfer of jurisdiction fordetermining the existence and magnitude ofdumping from the Department of Treasury tothe Department of Commerce. This seemingly

innocuous change was enormously consequen-tial.

While both agencies have been delegatedroles in the administration of U.S. economic pol-icy, the prevailing institutional objectives divergein many respects. Whereas Treasury’s mission isone of commitment to the broader macroeco-nomic well-being of the United States, thebenchmarks of success at Commerce have beenaligned historically with firm-specific outcomes.Whereas Treasury officials would be moreinclined to regard import duties resulting indeadweight economic loss as anathema to theirmission, Commerce officials are expected to pro-mote the narrow interests of domestic producers.Indeed, the mission of Import Administration,the agency within the Department of Com-merce tasked with antidumping oversight, is to“safeguard American industries and jobs againstunfair trade practices.”30 Import Administrationworks to “enforce effectively the U.S. unfair tradelaws (i.e., the antidumping and countervailingduty laws) and to develop and implement otherpolicies and programs aimed at countering for-eign unfair trade practices.”31 In that description,one can detect that Import Administration fan-cies itself responsible for writing, adjudicating,and enforcing the antidumping rules, all ofwhich should comport with a design to protectU.S. industry.

In the words of Ronald Cass, former vicechairman of the U.S. International Trade Com-mission, Commerce Department officials “seethemselves as advocates for domestic business.”32

And the advisory committee system establishedafter the 1974 Act provided private-sector stake-holders with direct channels of access to staff-level decisionmakers throughout the CommerceDepartment’s international trade bureaucracy.

A 1994 Congressional Budget Office studyof GATT and U.S. trade policy included com-mentary on the shift of antidumping jurisdic-tion from Treasury to Commerce:

The move reflected a Congressionaldesire for more zealous enforcement ofthe AD/CVD [antidumping/counter-vailing duties] laws and for less concernabout their being used in a protection-

10

ImportAdministration

fancies itselfresponsible for writing,

adjudicating,and enforcing

antidumping rules.

ist manner. Its significance goes beyondthe difference in the institutional sym-pathies. One of [Commerce]’s func-tions is to serve as an advocate for U.S.firms. Thus, the move placed responsi-bility for deciding AD/CVD cases inthe hands of an advocate of U.S. partiesto cases.33

The methodologies and procedures adopt-ed by the Commerce Department in adminis-tering antidumping have led many observers toquestion the agency’s impartiality. Some of theCommerce Department’s curious method-ological techniques and their margin-inflatingeffects have been documented in previous Catopublications.

In addition to the cost test and the use ofconstructed value (discussed earlier), practicessuch as the arm’s-length test, zeroing, the asym-metric treatment of indirect selling expenses as adeduction from price, and the deduction of prof-it from U.S. prices in so-called constructedexport-price transactions are some of the mostobvious methodological distortions that pro-duce egregiously skewed results.34

The table of counterfactuals from the 2002Cato study cited above demonstrates that thosevarious methodological distortions significantlyimpact the bottom-line result. For example, hadzeroing been precluded from the CommerceDepartment’s calculation methodology in the18 actual cases reviewed, the dumping marginswould have been 43.6 percent lower.35 Had theasymmetric treatment of indirect selling expens-es been disallowed, margins would have been9.1 percent lower.36

Of course, the standard methodological dis-tortions employed as a matter of course in thecalculation of antidumping margins at theCommerce Department do not cover all of thechannels through which agency bias affectsantidumping policy. The Commerce Depart-ment is granted enormous discretion in itsadministration of the law, and often its actionsare deemed by the courts to run afoul of the law.

As presented in a 2005 Cato study about theexercise of discretion at the CommerceDepartment, “IA [Import Administration] rou-

tinely exploits gray areas in the law to favor thedomestic interests that seek protection—and,according to the verdicts of U.S. courts, some-times violates the law in the process. In the 18-month period ending in June 2005, IA pub-lished 19 antidumping redeterminations pur-suant to court orders to revise its assumptions orcalculations to become compliant with the law.In 14 of those redeterminations, the revisedantidumping rates were lower than those origi-nally calculated.”37

According to interviews with former con-gressional and Commerce Department staffconducted by the Office of the InspectorGeneral:

This bias is illustrated by the actions ofcareer Commerce Department officialsthrough whom must pass all Depart-ment of Commerce [antidumping]determinations in steel cases. Themembers and staff of the congressionalSteel Caucus meet with them regularlyto discuss ongoing antidumping andcountervailing duty proceedings pend-ing before the Department of Com-merce. At some of these meetings,these officials have shared advance draftinvestigation results with the congres-sional Steel Caucus well before theywere announced in final form, allowingthe Steel Caucus to “comment” onthem. Time and again high level offi-cials within the agency have exertedpressure on lower level Department ofCommerce staff conducting investiga-tions of foreign steel producers to reruncalculations and alter methodologies,resulting in increased AD/CVD tar-iffs.38

Commerce investigators endeavor to sub-stantiate allegations of dumping and measure itsextent through lengthy, English-language ques-tionnaires consisting of five sections and sevenappendices, which are mailed to selected foreigncompanies and their embassies in Washington.These voluminous questionnaires seek fulldescriptions of virtually every aspect of the com-

11

The CommerceDepartment isgranted enormousdiscretion in itsadministration ofthe law, and oftenits actions aredeemed by thecourts to run afoulof the law.

panies’ immediate and related business activities,supplemented with databases containing recordsof the costs of producing and selling all subjectmerchandise in the United States, the homemarket, and various third-country markets. Theprocess almost always requires that the foreignrespondents enlist the services of law and eco-nomics firms with expertise in the law’s admin-istration. Agency guidelines stipulate that for-eign firms under investigation must respondwithin 37 days (and that Section A of the ques-tionnaire is due within 21 days).39 If responses tothe questionnaire are late, incomplete, or other-wise deemed unsatisfactory, Commerce investi-gators are authorized by Congress to utilize the“best information available” in determiningwhether dumping has taken place. Since thebest information available to the CommerceDepartment is often culled from the allegationsput forth by domestic competitors in their peti-tions for antidumping measures, this procedureis often punitive and unfair. In an opinionaddressed to the Court of International Trade(the appellate court for U.S. antidumping cases),

Judge R. Kenton Musgrave labeled the practicea “predatory ‘gotcha’ policy.”40

The data do little to dispel perceptions ofbias. Between 1980 (when Commerce tookover the administration of antidumping fromTreasury) and the end of 2009, a total of 1,091antidumping cases were initiated in the UnitedStates.41 Antidumping orders were imposed in590 of those cases.42 Figure 1 (above) accountsfor the dispositions of those 1,091 cases.

What stands out is that in only 46 of those1,091 cases did the Commerce Departmentrender a negative finding of dumping at the pre-liminary or final stage of the investigation. Inother words, petitioners have had a 96 percentsuccess rate at the Commerce Department sincethat agency took over antidumping administra-tion from Treasury. What prevented more ofthose initiations from resulting in antidumpingorders was the International Trade Commis-sion, which issued negative injury determina-tions in 359 cases.

By contrast, during its oversight of thedumping determination, the Treasury Depart-

12

Petitioners havehad a 96 percent

success rate at theCommerce

Department sincethat agency took

over antidumpingadministrationfrom Treasury.

Figure 1

Disposition of 1,091 Antidumping Initiations, 1980−2009

ment was far less likely to render affirmativefindings. Between 1934 and 1954, a total of 146U.S. antidumping cases were initiated andTreasury rendered negative dumping findings inat least 90 of those 146.43

The Commerce Department’s standing asan impartial arbiter is further undermined by awell-documented tendency to calculate exces-sive dumping margins. Foreign producers areroutinely found to be selling at prices so farbelow fair value as to raise serious doubt aboutthe accuracy and fairness of the Department’smethodology. As noted by Bruce Blonigen andThomas Prusa, the Commerce Department

not only finds dumping, they almostalways find unbelievably large dumpingmargins. Any argument that [anti-dumping] law is designed to ensure “fairtrade” looks ridiculous when confrontedwith [the Commerce Department’s]margins. According to the statute, thedumping duties are designed to makethe dumped imports “fairly traded”imports. Yet, the average dumping mar-gin over the past decade is about 60 per-cent.44

Even under the most extreme market condi-tions (i.e., perfect monopoly in the exportingcountry, perfect competition in the U.S. domes-tic market), price undercutting of 60 percentwould be overkill for all but the most aggressiveof predators. But that is often the outcome of aprocess in which dozens of decisions are madeby bureaucrats exercising judgment and discre-tion that may be influenced by the underlyingagency mission.

As noted in a 2005 Cato study:

When IA [Import Administration]publishes its final determinations, itoften includes a list of the issues raisedby the parties concerning methodolo-gies employed, calculations made, anddecisions rendered, along with a sum-mary of each party’s arguments and anexplanation for its ultimate judgmenton each decision. In the 18-month

period ending in June 2005, IA issued38 final determinations in originalantidumping investigations, findingdumping in 36 of those 38 determina-tions. In 11 of those final determina-tions, there were 5 or fewer issues raisedby the parties; in 13 of those determi-nations, there were 10 to 20 issuesraised; and in 14 of those cases therewere 21 or more issues that IA wasrequired to adjudicate. A total of 650issues—an average of more than 17 percase—were presented and requiredadjudication by IA in those 38 finaldeterminations. Thus, dumping calcu-lation involves more than simply plug-ging objective numbers into a comput-er program. In many cases, IA’s judg-ment calls determine the numbersthemselves.45

Relaxing “Material Injury”Standards

With an affirmative finding of sales at lessthan fair value by the Commerce Departmentall but assured after 1980, the last best hope forforeign firms (and U.S. importers) facing thespecter of antidumping restrictions was to pre-vail on the “injury side” of the investigation. Ifthere was no reasonable indication that thedomestic industry was “materially injured orthreatened with material injury,” or that “theestablishment of an industry was materiallyretarded” by reason of less than fair valueimports, then an antidumping order could notbe imposed, regardless of what the CommerceDepartment decided with respect to its dump-ing determination.

During its 25-year oversight of the injurydetermination function (from 1954 through1979), the Tariff Commission rarely foundmaterial injury to have been caused by dumping.Out of 641 antidumping case initiations duringthis period, the Tariff Commission renderedaffirmative injury findings in 97 cases—15 per-cent of the time.46 Of those 97 affirmative find-ings, though, 73 occurred after 1970, a year in

13

The CommerceDepartment’sstanding as animpartial arbiter is further undermined by awell-documentedtendency to calculate excessivedumping margins.

which the Tariff Commission rendered a prece-dent decision to consider imports from morethan one country—instead of imports from asingle country, as had been previous practice—when determining whether less than fair valueimports caused or threatened material injury.

That decision seems to have sparked anincrease in the number of subsequent anti-dumping filings, as well as the number of affir-mative rulings, because domestic industry peti-tioners learned quickly that if exporters of thesame or similar products from multiple coun-tries were simultaneously the subject of anti-dumping investigations, the chances of obtain-ing an affirmative injury finding were improved.Through the analytical practice now known as“cumulation,” if multiple countries are subject toan antidumping investigation of the same sub-ject merchandise, and if exports from any one ofthose countries, or all in combination, are foundto be a cause of material injury, then all must beconsidered injurious and subject to an anti-dumping order.

Though the practice of cumulation had beendiscretionary in the years after the 1970 TariffCommission decision, the Trade and Tariff Actof 1984 codified the practice as a requirement ofthe ITC’s injury analysis. It became a matter oflaw that imports from countries that alone couldnot be demonstrated to be a cause of materialinjury could be subjected to antidumping duties.

This analytical approach would seem to fly inthe face of any legitimate rationale for applyingantidumping measures. Dumping is a firm- andcountry-specific phenomenon. The artificial,unfair advantages that antidumping measuresare presumed to offset are supposed to be mar-ket-specific and the product of particular foreigngovernment policies. If exports from a particularcountry are found to have no competitiveadvantage in the U.S. market, then there are noadvantages to offset with antidumping duties.

But after 1984, the rules changed in a waythat created incentives for bringing more casesagainst more countries. Countries from whichimports of subject merchandise accounted forless than 3 percent of total imports of subjectmerchandise were considered negligible andimmune from antidumping duties unless all of

the countries that fell under the 3 percentthreshold “cumulatively” accounted for morethan 7 percent of total imports of subject mer-chandise.

As was to be expected, the number ofantidumping initiations increased dramatically,as the number of targeted products held steady,suggesting that petitioners were aware of thebenefits of the cumulation requirements. The“clarifications” of the 1984 Act moved anti-dumping administration even further away fromits theoretical underpinnings and signaled toU.S. industry and its trade lawyers that the rulesof the game had once again changed in theirfavor.

Antidumping was born nearly a century ear-lier as a mechanism to preserve and defendcompetitive markets and ultimately to protectconsumers. The 1984 Act completely trans-formed antidumping into a tool to suppresscompetition in the name of protecting domesticproducers from injury regardless of whether for-eign producers benefited from any unfair advan-tages in their home markets.

Identifying the Determinantsof Antidumping Use

As the foregoing history suggests, U.S.antidumping policy evolved in such a way as tobecome more accessible and more rewarding toU.S. import-competing industries. That thevolume of antidumping filings and the fre-quency of affirmative rulings underwent a dra-matic and sustained increase in the 1980s sug-gests that domestic industry was responsive tothose changes.

The definition of dumping had beenexpanded, domestic industry had been invited toplay a greater role in antidumping policy formu-lation and oversight (through the advisory com-mittee system), and the U.S. government agencymost closely aligned with domestic producerinterests was made responsible for investigatingdumping allegations. With the legislativechanges of 1984, U.S. firms seeking protectionwere given strong incentives to file even moreantidumping petitions to increase their odds of

14

U.S. antidumpingpolicy evolved insuch a way as to

become more accessible and more

rewarding to U.S.import-competing

industries.

obtaining affirmative injury findings at the ITC. Figure 2 shows upticks in usage in years fol-

lowing changes to the law, which would seemto confirm those characterizations. But thereare probably other reasons for the increase inantidumping use over time. For example, it isreasonable to assume that the demand forantidumping protection would be inverselyrelated to the level of general tariff protectionafforded. In other words, firms in import-competing industries might be more inclinedto seek antidumping protection in response toreductions in general tariffs.

Indeed, with the exception of a trend reversalin the 1920s and 1930s, declining tariff rateshave been evident throughout the history of theantidumping law. The average tariff from 1890to 1894 was 48.4 percent; from 1894 to 1897, itwas 41.3 percent; and when America’s firstantidumping law entered into force in 1916, theaverage tariff had fallen to 30.7 percent.47 TheFordney–McCumber Law of 1922 and Smoot-Hawley Tariff Act of 1930 reversed that down-ward trend, pushing the average up to as high as

59.1 percent in 1932. On the eve of U.S. entryinto GATT in 1947 the average tariff stood at20.1 percent and has fallen substantially in theyears since. The average collected rate in 1974—when the “cost test” was introduced into theantidumping regime—was 7.8 percent. Whenantidumping oversight was shifted fromTreasury to Commerce in 1980, the average tar-iff on dutiable imports was 5.6 percent. The ratehad fallen to 5.5 percent in 1984, when cumula-tion was made a statutory requirement. And itstood at 4 percent in 2008.48

The figure reveals an upward trend in theoverall number of U.S. antidumping initia-tions. Given the administrative changes thattook place between 1974, 1980, and 1984, it isnot surprising that the number of initiationsincreased substantially during that period, orthat the number of affirmative injury findingsincreased. Another interesting feature of thedata is that the number of products subject toantidumping investigations has, to a far greaterdegree than either the overall number of initi-ations or the number of affirmative findings,

15

Figure 2

Number of U.S. Antidumping Initiations, 1947−2009

Cas

e In

itia

tions

per

Yea

r

remained fairly consistent over time. That issuggestive of an increasing propensity of par-ticular subsets of U.S. producers to bring morecases. In fact, more than half of all currentlyoutstanding U.S. antidumping and counter-vailing duty orders (156 of 303) are againstimported chemicals and steel products.49 Theconcentration of protection in these few indus-tries is likely a product of the ITC cumulationrules giving incentive to petitioners to namemore target countries, as well as evidence thatthese industries have learned the ropes ofantidumping administration and are no longerdeterred by uncertainty about the process.

In 2004 Douglas Irwin published an analysisof the various determinants of increased anti-dumping use. Using data gathered from U.S.government sources, Irwin conducted a series ofregression analyses seeking to disentangle andmeasure the impact of independent variables,including GDP growth rate, the unemploymentrate, the exchange rate, tariff rate changes, andlegal and administrative changes to the anti-dumping law on the number of antidumpingcases filed per year between 1947 and 2002—the dependent variable.50

Irwin’s analysis found that unemploymentand exchange-rate appreciations were signifi-cant and positively related to the number ofantidumping case filings. That would seem tomake sense, since material injury is more easilydemonstrated when workers are let go and whenthere is evidence of sales lost to imports (whichhas been historically more likely when the U.S.dollar is appreciating). Irwin found the averagerate of tariff protection to be significant andnegatively related to the number of filings (i.e.,falling Most Favored Nation tariffs lead toincreased petitions), which is intuitive consider-ing that import competition tends to increase insectors where protection has been reduced. And,Irwin found the legal and administrativechanges to have had a significant positive impacton dumping filings after 1984—a trend likelyattributable to mandatory cumulation, but alsosuggestive of a lagged impact of the CommerceDepartment’s takeover of dumping investiga-tions.51

Irwin’s analysis considered antidumping ini-tiations through 2002. Since then, the predom-inant targets of U.S. antidumping activity havebeen exporters in China. Nearly 39 percent of all

16

Since 2002 the predominant

targets of U.S.antidumping

activity have beenexporters in China.

Figure 3

U.S. Antidumping Initiations against China (% of Total AD Initiations), 2000−2009

antidumping initiations since 2003 have target-ed China, and that rate has been increasing onan annual basis, as Figure 3 indicates.52

Certainly, imports from China increasedsubstantially between 2003 and 2009—whichmight explain some of the increase, as Irwinfound the average rate of tariff protection to beinversely related to the number of petitions.

But are Chinese exporters more likely tobenefit from unfair competitive advantages athome than exporters in other countries, or couldit be that the increase in the number of casesagainst China is driven by the fact that the spe-cial nonmarket economy (NME) methodologyreserved for China and just a few other countriesis fertile ground for the exercise of discretion andthe escalation of dumping margins?53 Nonmar-ket economy methodology is probably animportant determinant of the rising number ofantidumping initiations against Chinese ex-porters because affirmative and high dumpingmargins are more likely to be calculated underthis approach.54

Likewise, the change in practice, announced

by the Commerce Department in 2007, to startapplying the countervailing duty law to Chineseexporters is also a likely factor in the rise in casesagainst China over the past couple of years.Since material injury (or the threat of materialinjury) is a necessary condition of both anti-dumping and countervailing duties, there is lit-tle to lose and a lot to gain for U.S. industries byfiling both kinds of petitions simultaneously.The added incentive for doing so in Chinesecases (as opposed to cases where the traditionalmarket economy methodology is employed) isthat the combined duties can be especially puni-tive. Under NME methodology, home-marketprices are disregarded entirely and “normalvalue” is determined by estimating all of theinputs necessary to produce the good (labor,material, energy, overhead, etc.) and then assign-ing an estimated value to those inputs based ontheir prices in another, similarly sized, similarlydeveloped economy—usually India, in Chinesecases. The dumping margin is then calculated asthe average difference between the surrogatescreated and the U.S. prices.

17

The application ofthe countervailingduty law to Chineseexporters is also alikely factor in therise in cases againstChina over the pastcouple of years.

Figure 4

U.S. Antidumping and Countervailing Duty Initiations against China, 2000−2009

As woefully inaccurate as the surrogate maybe at approximating the Chinese home-marketprice, it is in fact a subsidy-free price. The costbuild-up is based on market-oriented prices,and any subsidized prices or factors are elimi-nated from consideration. In other words, ifone accepts the NME methodology as valid,then the margin calculated is sufficient to rem-edy the dumping margin and the subsidy mar-gin. The difference between the two capturesthe full distortion. Nevertheless, the Com-merce Department still separately calculatesthe benefit of the alleged subsidy and applies aseparate countervailing duty on top of the anti-dumping duty, resulting in a double countingof the subsidy, thereby creating an extra incen-tive for domestic industry to file an antidump-ing petition to supplement its countervailingduty petition.

As Figure 4 demonstrates, there has been anincrease in antidumping initiations against Chinasince the Commerce Department changed itscountervailing duty policy in 2007. Indeed, for 22of the 23 countervailing duty cases initiatedagainst China between 2007 and 2009, a com-panion antidumping case was initiated simulta-neously.

Congress’s Turf

Irwin’s statistical findings confirm what isapparent to the naked eye from Figure 1 andFigure 2: legal and administrative changes af-fecting the antidumping law have been signifi-cant determinants of subsequent antidumpinguse. That finding is important for several rea-sons.

First, it provides a powerful rejoinder toassertions that the antidumping law needs to bemade more accessible and trade restrictive so asto close loopholes in the battle against unfairtrade. When the law’s ease of use and its effec-tiveness at producing protection are more signif-icant determinants of antidumping use than anyother factor, it is reasonable to conclude that thelaw need not be made more user friendly.Second, Congress should seriously considerreform of the antidumping law to rein in the

abuse that its permissive rules and Commerce’sfox-guarding-the-henhouse oversight have en-gendered. No longer can the global economy becharacterized as a competition between “our”producers and “their” producers. Numerous U.S.interests—retailers, manufacturers, consumers,logistics providers, financial service providers,and more—are adversely affected when anti-dumping measures are imposed. Yet, under thelaw, the costs imposed on these groups are noteven permitted to be considered by the admin-istering authorities when rendering decisionsabout the propriety of antidumping measures.

So far Congress has been loath to acknowl-edge this reality. Instead, Congress considersany presentation of the economywide costs ofantidumping, any discussion of antidumpingreform, or any indictment of U.S. antidumpingpractices from the WTO or other trade disputesettlement bodies to be an affront to its author-ity and an encroachment on its jurisdiction.

Efforts in 2002 by then−U.S. trade represen-tative Robert Zoellick to get antidumping ruleson the agenda of the Doha Round of multilat-eral trade negotiations (an absolute condition ofgetting the Round launched) were made notori-ously difficult by a Congress that insisted onlanguage that committed the USTR to “notweaken” the antidumping law. When Zoellickreturned from the meeting that ultimately pro-duced the Doha Declaration, he was summonedto Congress to explain why he had agreed toinclude negotiations on antidumping on theagenda, even though the language did not com-mit anybody to any reform.

In the debate over fast-track trade negotiat-ing authority in 2002, Congress strongly consid-ered unbundling negotiations over antidumpingfrom that authority, as proposed in the so-calledDayton-Craig Amendment. That rule wouldhave given the president the authority to bringback trade agreements for an up- or down-votewithout amendments, but with the exceptionthat those same fast-track rules would not applyto any changes to antidumping rules, whichwould require the full congressional treatment.

And to this day, there is continued resis-tance in Congress to implementing the find-ings of the WTO dispute settlement body over

18

It is reasonable toconclude that the

law need not bemade more

user friendly.

the issue of zeroing—the controversial methodfor calculating antidumping duties.55

One cannot help but acquire the impressionthat Congress and the Department of Com-merce are unwilling to relinquish what they con-sider a back channel for special favors to con-stituent interests that may find themselves inneed of assistance.

Emboldened by this institutional resistanceto curbing antidumping abuse, import-compet-ing interests have taken the initiative to push forsome new reforms of their own. However, thosereforms would make antidumping measureseasier to obtain and more difficult for foreignexporters and their U.S. customers to escape.

Based on a solicitation of comments from“the public,” the Commerce Departmentrecently announced its intention to strengthenthe enforcement of U.S. trade remedies laws.56

Among the various reforms under considerationis one that would make it impossible for indi-vidual companies to get themselves out fromunder a dumping order even when they demon-strate that they are not dumping. Currently,individual companies from a foreign country canbe excused from an antidumping order bydemonstrating that they are not dumping for acertain period of time (usually three consecutiveyears). The new proposal would eliminate thatchannel and force all exporters from a givencountry to remain subject to the antidumpingorder for as long as it exists.

Another facially innocuous proposal thatexperts see as a cunning effort to makeantidumping regulations even more restrictivewould change the process by which theCommerce Department selects foreign respon-dents in antidumping investigations andreviews. Usually, Commerce chooses to investi-gate or review the producers accounting for thelargest volume of exports to the United States.But the proposal on the table would requireCommerce to use random sampling in selectingthe respondents, which would increase theresource burdens on smaller firms that might beunable to afford the legal and economic repre-sentation that antidumping proceedings entail.Accordingly, Commerce would likely see lesscompliance with the arduous requirements of

the proceedings and would likely find itselfresorting to “Facts Available” findings withgreater frequently. Facts Available findings,which usually substitute figures alleged in thedomestic industry’s petition for the exporter’sactual data, almost always produce higherdumping margins for the company in ques-tion—and for all the other companies that werenot individually investigated—because theirexports are subject to the average duty calculat-ed for all investigated companies.

Economists Michael Moore of GeorgeWashington University and Thomas Prusa ofRutgers University summarized the problemwith this recent example:

Suppose three large firms account for90 percent of the subject imports while20 others account for the remaining 10 percent. Under current rules theDepartment of Commerce samples thebig three firms. Under the proposedrules, it can instead sample the pricingby firms with, say, 1 percent of theimports. What is the logic of focusingon such small firms? The reason is thatsmall firms may not have the resourcesto spend the $2 million (or more)required to participate in a proceeding.Failure to respond to all questionsallows the Department of Commerceto use “facts available” (i.e., domesticfirms’ allegations) when computingmargins. Blonigen (2006) finds thatusing “facts available” increases theaverage computed margin by 30 per-centage points.”57

If these kinds of changes are implemented,there should be little doubt of the impact onthe number of case initiations. As has beendemonstrated, episodes of relaxation of stan-dards and rules over the years have been fol-lowed by increased usage. And with any causallink between a foreign firm’s prices in theUnited States and anti-competitive practices inthe home market entirely severed by the pro-gressive lowering of evidentiary standards andthe expansion of the definition of dumping,

19

Episodes of relaxation of standards and rulesover the years havebeen followed byincreased usage ofthe antidumpinglaw.

antidumping measures will be even more dis-ruptive of trade and progress, benefiting only aselect few politicians and well-connected firms.

Conclusion

U.S. antidumping law no longer seeks toadvance the pro-competition objectives of itsoriginal charter. What began as an outgrowth ofantitrust intended to discipline predatory pric-ing—and, therefore, to safeguard competition—was gradually transformed into a mishmash ofrules and discretion that enables inefficient,uncompetitive behavior on the part of domesticfirms at great cost to other domestic entities.

Antidumping has done little, if anything, tofocus the spotlight on the alleged sources ofunfairness that presumably give rise to cross-border price discrimination. By systemicallydisregarding that kind of inquiry, antidump-ing’s overseers have blurred any distinctionsbetween possibly objectionable price discrimi-nation and the kind that is the product of profit-maximizing decisionmaking. Further-more, market access barriers and transportationcosts have fallen precipitously since the estab-lishment of antidumping policy, removing theimpediments to arbitrage necessary for cross-border price discrimination to occur withoutundermining the market power of the foreignproducer.

Although the real-world economic rationalefor the antidumping status quo is virtuallyextinct, political support for this favor-dolingmachine remains strong and bipartisan. Thatantidumping did a lousy job living up to itsrhetoric did not stop policymakers from ex-panding the definition of dumping and loweringthe evidentiary requirements necessary forindustry to obtain antidumping protection. Thenumerous legal and administrative changes,such as enabling the attribution of materialinjury to competitors in multiple countries, anddisregarding home-market sales at prices belowthe full cost of production, further underminedthe rationale for antidumping and blurred itsdistinction from run-of-the-mill, everyday pro-tectionism.

This paper has presented a historicalaccount of the events punctuating antidump-ing’s metamorphosis and provided some num-bers explaining the increasing resort to anti-dumping. The data seem to support thehypothesis that increasing demand for anti-dumping protection corresponds not withincreased dumping, as defined by economists,but rather with the increased flexibility ofantidumping, as defined by politicians. Andongoing efforts to make antidumping evenmore flexible are likely to induce more abuse—at great expense to the U.S. and global econ-omies.

NotesThe author owes a debt of gratitude to Todd Fox,whose excellent December 2009 graduate paper,“Changing the Definition of Fair: The EconomicHistory of Antidumping in the United States,”submitted to Johns Hopkins School of AdvancedInternational Studies, was the inspiration for thispaper and, indeed, the model for its structure.

1. “Basic Concepts and Principles of the TradeRemedy Rules,” communication from the UnitedStates to the WTO Negotiating Group on Rules,TN/RL/W/27, October 22, 2002.

2. The antidumping statute is codified at 19U.S.C. §§ 1673 1677n. The DOC’s antidumpingregulations may be found at 19 C.F.R. § 351.

3. In cases involving so-called nonmarket econo-mies, the DOC calculates constructed value, notwith the foreign producer’s own cost data, but withdata from “surrogate” market economies. For acomprehensive treatment of nonmarket economymethodology, see Daniel J. Ikenson, “NonmarketNonsense: U.S. Antidumping Policy towardChina,” Cato Trade Briefing Paper no. 22, March 7,2005.

4. The antidumping regime’s failure to discernthe causes of price discrimination represents afatal blow to its legitimacy. But beyond that, themethodologies employed to determine and calcu-late the extent of price discrimination (i.e.,“dumping margins”) are egregiously tilted infavor of finding larger margins. For a comprehen-sive critique of those methodologies, see BrinkLindsey and Dan Ikenson, “Antidumping 101:The Devilish Details of ‘Unfair Trade’ Law,” CatoTrade Policy Analysis no. 20, November 21, 2002.

5. Michael P. Galloway, Bruce A. Blonigen, and

20

Although the real-world economicrationale for the

antidumping statusquo is virtually

extinct, politicalsupport for this

favor-dolingmachine remains

strong and bipartisan.

Joseph E. Flynn, “Welfare Costs of the U.S. Anti-dumping and Countervailing Duty Laws, Journalof International Economics 49 (1999): 236.

6. Douglas Irwin, “The Rise of U.S. AntidumpingActions in Historical Perspective,” NBER Work-ing Paper 10582, June 2004, p. 4.

7. Ian Wooton and Maurizio Zanardi, “Trade andCompetition Policy: Anti-Dumping Versus Anti-Trust,” Department of Economics, University ofGlasgow, Working Papers 2002_6, October 2002,p. 2.

8. Irwin, p. 3.

9. J. Michael Finger, Antidumping: How It Works andWho Gets Hurt (Ann Arbor: University of MichiganPress, 1993), p. 21.

10. Ibid., p. 21.

11. Ibid., p. 14.

12. Ironically, U.S. Steel would emerge in subse-quent decades as one of the prime users of anti-dumping.

13. Finger, p. 16.

14. Ibid., p. 19.

15. Irwin, p. 4.

16. Joshua Bernhardt, quoting the United StatesTariff Commission Report of 1919, in The TariffCommission, Its History, Activities and Organization(New York: D. Appleton and Company, 1922), p.35, http://books.google.com/books?id=M-9AAAAAIAAJ&printsec=frontcover&dq=Joshua+Bernhardt+The+Tariff+Commission,+Its+History,+Activities+and+Organization+1922&source=bl&ots=_XIb-PzaK0&sig=1d6792tLq4y4_46gLCSx81KLWGY&hl=en&ei=ZQn9TJP8L8H7lweNpLyMBQ&sa=X&oi=book_result&ct=result&resnum=1&ved=0CBgQ6AEwAA#v=onepage&q&f=false.

17. Congressional Budget Office, The Evolution ofU.S. Laws: An Economic Perspective, 1994, p. 21.

18. Finger, p. 21.

19. Ibid., p. 22.

20. Irwin, p. 6.

21. World Trade Organization, Article VI, GATT1994.

22. Finger, p. 26.

23. Ibid.

24. Per the Final Act of the 1964–1967 Trade Con-

ference Section B, Article 3: “In order to establishwhether dumped imports have caused injury, allother factors which, individually or in combina-tion, may be adversely affecting the industry shallbe examined, for example: the volume and pricesof undumped imports of the product in question,competition between the domestic producersthemselves, contraction in demand due to substi-tution of other products or to changes in con-sumer tastes.”

25. Irwin, p. 7.

26. For a more detailed accounting of domesticindustry’s lobbying of Congress on this issue, seeWilliam H. Barringer and Kenneth J. Pierce, Payingthe Price for Big Steel (Washington: AmericanInstitute for International Steel, 2000), p. 75.

27. Barringer and Pierce, 75n93.

28. Brink Lindsey and Dan Ikenson, “Antidumping101: The Devilish Details of ‘Unfair Trade’ Law,”Cato Trade Policy Analysis no. 20, November 21,2002, p. 14.

29. Ibid.

30. Import Administration website, http://trade.gov/ia/.

31. Ibid.

32. Michael Anthony Lawrence, “Bias in the Inter-national Trade Administration: The Need forImpartial Decisionmakers in United States Anti-dumping Proceedings,” Case Western Reserve Journalof International Law, Winter 1994, p. 3.

33. Congressional Budget Office, “How the GATTAffects U.S. Antidumping and Countervailing-Duty Policy,” September 1994.

34. For a detailed discussion of several of thesemethodological distortions and their impact, seeDaniel J. Ikenson, “Abuse of Discretion: Time toFix the Administration of the U.S. AntidumpingLaw,” Cato Trade Policy Analysis no. 31, October 6,2005; and Brink Lindsey and Dan Ikenson,“Antidumping 101: The Devilish Details of ‘UnfairTrade’ Law,” Cato Trade Policy Analysis no. 20,November 21, 2002.

35. Lindsey and Ikenson, p. 14.

36. Ibid.

37. Daniel J. Ikenson, “Abuse of Discretion: Timeto Fix the Administration of the U.S. Antidump-ing Law,” Trade Policy Analysis no. 31, October 6,2005, p. 1.

38. Barringer and Pierce, p. 81.

21

39. U.S. Department of Commerce, Office of AD/CVD Enforcement.

40. Lawrence, p. 3.

41. Compiled from databases available on theImport Administration website at http://ia.ita.doc.gov/stats/iastats1.html.

42. The 599 antidumping measures include 25suspension agreements that are effectively settle-ments involving the agreement of the foreignexporters to sell above a certain price in theUnited States.

43. Cited in Irwin at p. 24, U.S. House of Represen-tatives, Committee on Ways and Means, “Hearingson Amendments to the Antidumping Act of 1921,as Amended,” 85th Congress, 1st Sess., 1957, p. 15.“At least” 90, and probably more, negative determi-nations were rendered because the data included inthe cited report aggregate “de minimis” findings ofdumping with “complaints withdrawn” and otherreasons for non-affirmative findings. But de min-imis findings are akin to negative finding of dump-ing and should have been counted as negatives.

44. Bruce A. Blonigen and Thomas A. Prusa, “Anti-dumping,” in E. K. Choi and J. Harrigan, Handbookof International Trade (Oxford, UK and Cambridge,MA: Blackwell Publishers, 2003), p. 267.

45. Ikenson, October 2005, p. 3.

46. Tallied from Irwin, Appendix, p. 22.

47. Calculated as the ratio of duties collected overthe value of imports for consumption.

48. United States International Trade Commis-sion, Value of U.S. Imports for Consumption, DutiesCollected, and Ratio of Duties to Values 1891–2008,ITC Statistical Services Division, April 2009.

49. Compiled from AD/CVD statistics availableon the website of the U.S. International TradeCommission, http://info.usitc.gov/oinv/sunset.nsf/0a915ada53e192cd8525661a0073de7d/96da

f5a6c0c5290985256a0a004dee7d/$FILE/orders%20August%2012%202010.xls.

50. The legal and administrative changes testedwere those made pursuant to the Trade Reform Actof 1974 (the inclusion of the “cost test”), the 1980shift of oversight from Treasury to Commerce, andthe codification of “cumulation” pursuant to theTrade Agreements Act of 1984. Irwin used dummyvariables corresponding to the first full years of thenew policies: 1975, 1980, and 1985, respectively.

51. Irwin, p. 25.

52. Compiled from AD/CVD statistics available onthe website of the U.S. International Trade Com-mission, http://info.usitc.gov/oinv/sunset.nsf/0a915ada53e192cd8525661a0073de7d/96daf5a6c0c5290985256a0a004dee7d/$FILE/orders%20August%2012%202010.xls.

53. For a detailed discussion of non-market econo-my methodology, see Daniel J. Ikenson, “Nonmar-ket Nonsense: U.S. Antidumping Policy towardChina,” Cato Trade Briefing Paper no. 22, March 7,2005.

54. Ibid., p. 6.

55. For a detailed discussion of this issue, seeDaniel J. Ikenson, “Zeroing In: Antidumping’sFlawed Methodology under Fire,” Cato Free TradeBulletin no. 11, April 27, 2004.

56. U.S. Department of Commerce, Press Re-lease, August 26, 2010, “Obama AdministrationStrengthens Enforcement of U.S. Trade Laws inSupport of President’s National Export Initia-tive,” http://www.commerce.gov/news/press-releases/2010/08/26/obama-administration-strengthens-enforcement-us-trade-laws-support-pr.

57. Michael Moore and Thomas Prusa, “StraightOut of Mad Mend: US Markets Import Protectionas Export Prompotion,” Vox Commentary, No-vember 8, 2010, http://www.voxeu.org/index.php?q=node/5751.

22

Trade Policy Analysis Papers from the Cato Institute

“The U.S. Generalized System of Preferences: Helping the Poor, But at What Price?” by Sallie James (no. 43; November 16,2010)

“Made on Earth: How Global Economic Integration Renders Trade Policy Obsolete” by Daniel Ikenson (no. 42; December2, 2009)

“A Harsh Climate for Trade: How Climate Change Proposals Threaten Global Commerce” by Sallie James (no. 41;September 9, 2009)

“Restriction or Legalization? Measuring the Economic Benefits of Immigration Reform” by Peter B. Dixon and Maureen T.Rimmer (no. 40; August 13, 2009)

“Audaciously Hopeful: How President Obama Can Help Restore the Pro-Trade Consensus” by Daniel Ikenson and ScottLincicome (no. 39; April 28, 2009)

“A Service to the Economy: Removing Barriers to ‘Invisible Trade’” by Sallie James (no. 38; February 4, 2009)

“While Doha Sleeps: Securing Economic Growth through Trade Facilitation” by Daniel Ikenson (no. 37; June 17, 2008)

“Trading Up: How Expanding Trade Has Delivered Better Jobs and Higher Living Standards for American Workers” byDaniel Griswold (no. 36; October 25, 2007)

“Thriving in a Global Economy: The Truth about U.S. Manufacturing and Trade” by Daniel Ikenson (no. 35; August 28,2007)

“Freeing the Farm: A Farm Bill for All Americans” by Sallie James and Daniel Griswold (no. 34; April 16, 2007)

“Leading the Way: How U.S. Trade Policy Can Overcome Doha’s Failings” by Daniel Ikenson (no. 33; June 19, 2006)

“Boxed In: Conflicts between U.S. Farm Policies and WTO Obligations” by Daniel A. Sumner (no. 32; December 5, 2005)

“Abuse of Discretion: Time to Fix the Administration of the U.S. Antidumping Law” by Daniel Ikenson (no. 31; October 6,2005)

“Ripe for Reform: Six Good Reasons to Reduce U.S. Farm Subsidies and Trade Barriers” by Daniel Griswold, StephenSlivinski, and Christopher Preble (no. 30; September 14, 2005)

“Backfire at the Border: Why Enforcement without Legalization Cannot Stop Illegal Immigration” by Douglas S. Massey(no. 29; June 13, 2005)

“Free Trade, Free Markets: Rating the 108th Congress” by Daniel Griswold (no. 28; March 16, 2005)

“Protection without Protectionism: Reconciling Trade and Homeland Security” by Aaron Lukas (no. 27; April 8, 2004)

“Trading Tyranny for Freedom: How Open Markets Till the Soil for Democracy” by Daniel T. Griswold (no. 26; January 6,2004)

“Threadbare Excuses: The Textile Industry’s Campaign to Preserve Import Restraints” by Dan Ikenson (no. 25; October 15, 2003)

Board of Advisers

James Bacchus Greenberg Traurig LLP

Jagdish BhagwatiColumbia University

Donald J. BoudreauxGeorge Mason University

Douglas A. IrwinDartmouth College

José PiñeraInternational Center forPension Reform

Russell RobertsGeorge Mason University

Razeen SallyLondon School ofEconomics

George P. ShultzHoover Institution

Clayton YeutterFormer U.S. TradeRepresentative

The mission of the Cato Institute’s Center for Trade Policy Studies is to increase publicunderstanding of the benefits of free trade and the costs of protectionism. The center

publishes briefing papers, policy analyses, and books and hosts frequent policy forums andconferences on the full range of trade policy issues.

Scholars at the Cato trade policy center recognize that open markets mean wider choicesand lower prices for businesses and consumers, as well as more vigorous competition thatencourages greater productivity and innovation. Those benefits are available to any countrythat adopts free-trade policies; they are not contingent upon “fair trade” or a “level playingfield” in other countries. Moreover, the case for free trade goes beyond economic efficiency.The freedom to trade is a basic human liberty, and its exercise across political borders unitespeople in peaceful cooperation and mutual prosperity.

The center is part of the Cato Institute, an independent policy research organization inWashington, D.C. The Cato Institute pursues a broad-based research program rooted in thetraditional American principles of individual liberty and limited government.

For more information on the Center for Trade Policy Studies,visit www.freetrade.org.

Other Trade Studies from the Cato Institute

“The U.S. Generalized System of Preferences: Helping the Poor, But at What Price?” by SallieJames, Trade Policy Analysis no. 43 (November 16, 2010)

“A Free Trade Agreement with South Korea Would Promote Both Prosperity and Security”by Doug Bandow, Trade Briefing Paper no. 31 (October 20, 2010)

“The Miscellaneous Tariff Bill: A Blueprint for Future Trade Expansion” by Daniel Griswold,Trade Briefing Paper no. 30 (September 9, 2010)

“Manufacturing Discord: Growing Tensions Threaten the U.S.-China EconomicRelationship” by Daniel J. Ikenson, Trade Briefing Paper no. 29 (May 4, 2010)

“Made on Earth: How Global Economic Integration Renders Trade Policy Obsolete” byDaniel Ikenson, Trade Policy Analysis no. 42 (December 2, 2009)

“A Harsh Climate for Trade: How Climate Change Proposals Threaten Global Commerce”by Sallie James, Trade Policy Analysis no. 41 (September 9, 2009)

“Restriction or Legalization? Measuring the Economic Benefits of Immigration Reform” byPeter B. Dixon and Maureen T. Rimmer, Trade Policy Analysis no. 40 (August 13, 2009)

CENTER FOR TRADE POLICY STUDIES

Nothing in Trade Policy Analysis should be construed as necessarily reflecting the views of theCenter for Trade Policy Studies or the Cato Institute or as an attempt to aid or hinder the pas-sage of any bill before Congress. Contact the Cato Institute for reprint permission. Additionalcopies of Trade Policy Analysis studies are $6 each ($3 for five or more). To order, contact theCato Institute, 1000 Massachusetts Avenue, N.W., Washington, D.C. 20001. (202) 842-0200, fax (202) 842-3490, www.cato.org.