Dumping on free trade: the U.S. import trade laws.

1. Introduction

As import tariffs have fallen across the world, nontariff barriers have become more important impediments to trade. The use of unfair trade laws, in particular, has increased dramatically over the past two decades - undoubtedly much more dramatically than any increase in unfair trade itself. In the United States, for example, 233 antidumping cases were investigated between 1968 and 1978, whereas almost twice that number (451 cases) were investigated between 1980 and 1989.(1) Unfair trade laws are also being used more frequently abroad: Figure 1 shows the number of antidumping cases filed by non-U.S. GATT participants since 1980 (note the apparent counter-cyclical pattern of the series).(2)

In December 1994, President Clinton signed the Uruguay Round implementing legislation - the culmination of the eighth round of international negotiations under the General Agreement on Tariffs and Trade (GATT). The conclusion of the Uruguay Round seems an appropriate occasion to review the policy goals of our import trade statutes,(3) to assess how well our current laws achieve those objectives, and to explore future courses of action. This paper is therefore organized into four sections: Section 2 discusses the appropriate policy objectives of our import trade laws, section 3 describes the operation of our pre-Uruguay Round statutes, section 4 explores some of the changes effected by the Uruguay Round, and section 5 examines possible directions for future policies.(4)

2. What Are the Proper Policy Objectives of Import Trade Laws?

A well-designed set of import trade laws would address at least three crucial economic policy concerns: dumping, foreign export subsidies, and import surges.(5) This section will discuss such concerns and their impact on national economic welfare.

Dumping

Dumping is conventionally defined as a type of international price discrimination: the sale of goods within the United States at a price lower than in some foreign market.(6) Ironically, viewed from a static perspective, laws proscribing such international price discrimination prevent behavior beneficial to the U.S. and condone behavior detrimental to the U.S. Consider the example of a monopolistic foreign firm. If there are barriers to trade that facilitate price discrimination, our static welfare is maximized if the foreign firm discriminates when our demand is relatively elastic and does not discriminate when our demand is relatively inelastic. (Under conventional assumptions, the price charged to U.S. consumers would be lower than a single pooled price if our demand is more elastic than foreign demand. It would be higher than a pooled price if our demand is less elastic than foreign demand.) But antidumping laws prohibit price discrimination in the first case and not in the second - exactly the opposite of what our optimal policy should be! Even in more complicated static cases (for example, oligopolistic competition) with domestic producers, dumping is still frequently welfare-improving.(7)

From a static perspective, then, dumping by foreign firms frequently seems to make us better off. But there are two reasons to be wary of the predictions of the static perspective and thus two legitimate reasons for policymakers to worry about dumping: predatory pricing and new trade theory effects.

Predatory Pricing

Economists have long recognized the possible welfare losses caused by predatory pricing, in which a firm intentionally drives down the market price in order to force other firms out of the market.(8) Compared to the counterfactual of no predatory behavior, consumer surplus is higher during the predation period (since the price is lower) but then lower in the subsequent period (since, if the predation is successful, market power and the price are higher). After discounting and integrating appropriately, it is certainly possible that consumer surplus will be reduced by predation. Does economic theory offer policymakers any guidance in assessing the likelihood of such losses?

The theory of predatory pricing has evolved substantially over the past decade.(9) The older Chicago view is that predatory pricing is unlikely to occur, largely because in most situations it is implausible for firms to expect such a strategy to generate positive (discounted) profits. According to this view, the predation period may have to be relatively prolonged and costly. Entry during the later period, furthermore, is likely to dissipate any incipient rents. The implication is that predatory pricing is rarely a profitable activity and therefore should rarely be observed. New developments in industrial organization theory, however, suggest a somewhat less sanguine perspective: They posit a variety of ways in which predators can prevent entry and thus protect the rents accruing during the second stage. The strategic entry deterrence literature, for example, presents models with irreversible investment, in which firms can deter entry by building excess capacity or investing in advertising. Such activities can be expensive, but they may make predatory pricing a profitable strategy by protecting the second period rents. Alternatively, predatory pricing itself can be a deterrent to entry: Such behavior may create a reputation for craziness and irrational behavior, and this reputation may dissuade potential rivals from entering in the later period. Either way, a predatory pricing strategy may be more attractive to firms than the Chicago view would suggest.

If predatory pricing actually occurs, a foreign predator may reduce domestic welfare more than a domestic predator ceteris paribus for at least two reasons.(10) First, a predatory strategy will be adopted only if it generates expected discounted profits in excess of the cost of exercising the option to undertake the predation.(11) Thus, monopolization of a domestic market by a foreign producer creates an expected domestic welfare loss both from the deadweight loss (as in purely domestic cases) and from the expected present discounted value of the differential profits accruing from the strategy, which represent an expected transfer of rents from domestic consumers to the foreign producer. Second, labor is usually thought to be relatively mobile within a country - especially in the United States - but relatively immobile between countries. If workers in the firm that is attacked have skills specific to the industry, the transfer of production to the predator firm generates more domestic dislocation (in terms of unemployment mismatch) if the expanded production occurs abroad rather than at home. This seems more likely if the predator is a foreign firm rather than a domestic one.

Predatory dumping may thus be a legitimate policy concern. But it is important to note that dumping, as conventionally defined, is neither necessary nor sufficient for predation in practice. Predation can occur even if the price is higher in the U.S. than abroad (if the U.S. price, for example, would have been even higher, relative to the foreign price, in the absence of the predatory behavior). And price discrimination across national boundaries can be nonpredatory (if, for example, elasticities of demand for a firm's product vary across countries). To foreshadow the discussion in section 3, I should also stress that a U.S. price below the foreign exporter's average cost is neither necessary nor sufficient for predation. Predation is possible even if the price is above the predator's average cost (if, for example, the price is below both the prey's marginal cost and the predator's short-run profit-maximizing price(12)). And prices below average cost - or even below short-run marginal cost - are not prima facie evidence of predation (see section 3).

New Trade Theory Effects

The new trade theory suggests that there may be a limited number of cases, involving some sort of nonconvexity and imperfect competition, in which protection of the home market affords a firm first-mover advantages, facilitates dynamic learning effects, or generates other benefits.(13) I should note, however, that it is not obvious that outright protection of the home market, as opposed to subsidization, actually bestows any new trade theory advantages on domestic firms. Under a protectionist policy, the domestic firm obtains a larger share of a smaller market (assuming a downward-sloping domestic demand curve). Whether or not protection produces new trade theory benefits for the domestic firm thus depends on the specific form of the relevant new trade theory effect, the effectiveness of the protection in excluding foreign firms, and product demand parameters (see Brander and Spencer 1984; Krugman 1984).

Protectionism may not produce new trade theory benefits in all industries, but the potential gains in some industries raise complex policy questions. In particular, in a new trade theory world, the strong (conventional) form of predation must be supplemented by a weak form of predation, in which the foreigner's objective is to acquire or appropriate rents from domestic firms rather than to drive them out of business.(14) The weak form of predation suggests that antidumping laws could potentially boost national welfare either by allowing domestic firms to gain new trade theory advantages or by preventing foreign firms from obtaining them.

Foreign Export Subsidies

In the majority of cases, it is in our national interest to welcome subsidized exports from other countries. Such subsidized exports most frequently occur in industries in which we are net importers so, at least from a partial equilibrium analysis of the market for the imported good, the consumer surplus gain from any price decline exceeds the producer surplus loss.(15) Thus, economists often argue that we should be grateful for the gift, perhaps even sending a "thank-you note," as one trade scholar glibly put it (Jackson 1989, p. 251; also see Dixit 1987). But there are two reasons to temper our general enthusiasm for subsidized imports: the new trade theory (again) and global welfare maximization.

New Trade Theory Effects

The new trade theory suggests that there may be some cases in which export subsidization could raise national welfare and reduce welfare in the importing country (Brander and Spencer 1985). In this case, countervailing duty laws (which countervail subsidies that promote exports) can be useful in preventing foreign firms from gaining a first-mover advantage. Similarly, multilateral disciplines and foreign countervailing duty laws can mitigate the incentives for us to try to gain such an advantage.

Global Welfare Effects

Even within the conventional paradigm of perfectly competitive markets, the existence of countervailing duty laws could be beneficial from a global welfare perspective. Such laws attempt to counteract the effects of export subsidies and thus may dissuade misguided policymakers from granting the subsidies in the first place.(16) Put another way, subsidies that distort the trade flows consistent with comparative advantage reduce global welfare (again, assuming a perfectly competitive world). Countervailing duty laws may be useful in discouraging such behavior. The problem is that the U.S. has to threaten to shoot itself in the foot in order to ensure that others do not shoot themselves in their respective feet. We can only hope that the threat is credible enough that we will not have to execute it (and hence ourselves).

A final subtlety concerning countervailing duties is the treatment of externalities. Global welfare may be boosted by government subsidies that encourage activities with positive externalities (e.g., research and development [R&D] and other activities). And to the extent that the externalities cross national boundaries, our national welfare may be raised by foreign government subsidies. So there is some motivation for exempting R&D and other similar subsidies from the countervailing duty laws. At the same time, however, it is often difficult in practice to distinguish a production subsidy from an R&D subsidy. This is the rationale for the treatment of R&D subsidies in the Uruguay Round agreement (see section 4): They are protected from countervailing duties, i.e., are green-lighted, but there are limits on what qualifies as an R&D subsidy.

Import Surges

With large adjustment costs and some form of market failure (for example, capital market imperfections), it may be in the national interest for the government to provide a means of smoothing adjustment to very large but temporary trade shocks, basically as a type of insurance policy, even though such shocks do not represent unfair trade.(17) The problem with this kind of policy is that it is not possible to discern whether a given shock is temporary or permanent until well after the shock has hit. Since policies that impede adjustment to permanent shocks are likely to be welfare-reducing, the optimal market surge policy must compare the (potential) benefits of providing insurance against temporary surges to the (probable) costs of preventing or delaying required adjustments to permanent shocks.

Whether or not the market surge laws provide a direct economic net benefit to the country, they can be supported on political economy grounds. Some level of insurance seems to be necessary to make a free trade regime politically acceptable, and thus a market surge protection law may be required to obtain the overall benefits of a liberal trading regime. The GATT recognizes this reality in Article XIX, which allows safeguards to be imposed if unanticipated increases in imports inflict serious injury on a domestic industry.

3. The Pre-Uruguay Round Trade Laws

The U.S. has an elaborate and comprehensive system of trade laws, many of which are referred to by their statutory numbers. Sections 201 and 406 address market surges? Section 232 regulates imports that threaten national security, Section 301 deals with foreign trade barriers and violations of trade agreements, Section 337 addresses unlawful marketing of imports (such as infringements of patents), Section 701 governs countervailing duties, and Section 731 is the antidumping statute.(19) My focus here will be the antidumping, countervailing duty, and market surge laws. I will also discuss Section 337 briefly.

Antidumping Laws

The current antidumping process has two parts: a dumping component and an injury component. Under the dumping part of the process, the Department of Commerce compares the price in the U.S. to a so-called foreign market value; the dumping margin is then simply the difference between the foreign market value and the U.S. price. Under the injury part, the International Trade Commission (ITC) determines whether the imports are causing or threatening material injury to the. U.S. industry. (This injury standard is substantially weaker than in the import surge laws.) Both dumping and injury must be found for final duties to be imposed, and relief must be granted if both findings are positive.

Unfortunately, our antidumping laws no longer have much to do with the prevention of the strong form of predatory pricing.(20) Our first antidumping statute - the Antidumping Act of 1916 - was clearly aimed at preventing international predatory pricing.(21) But since the 1921 Antidumping Act, it has not been necessary to demonstrate predatory intent or effect, so any international price discrimination (whether predatory or not) in which the U.S. price is lower than the foreign price has been proscribed.(22)

Since the 1974 trade act, the focus of the antidumping laws has shifted from preventing price discrimination to preventing sales below fully allocated average cost.(23) The mechanism driving this result is buried deep in the trade statutes. The law defines a hierarchy of methods for determining the foreign market value of a good: the home market index (the price at which the good is sold in the exporter's home market), the third country index (the price at which the good is sold in another of the exporter's foreign markets), and, only as a last resort, constructed value (an estimate of the exporter's average cost)? This hierarchy seems somewhat reasonable. But the law also stipulates that sales below total cost must be disregarded when calculating the first two indices if such sales have been made "over an extended period of time and in substantial quantifies, and are not at prices which permit recovery of all costs within a reasonable period of time in the normal course of trade" (Committee on Ways and Means 1993, p. 388). In practice, so many observations are excluded because of this provision that foreign market value must often be calculated using the constructed value index.(25)

Prohibiting sales below average total cost precludes many perfectly rational and nonpredatory activities.(26) For example, economists have long recognized that profit-maximizing firms may have sales below average total cost but above average variable cost in the short run. Furthermore, as Spence (1981) and others have shown, learning curves can induce firms to forward price at expected long-run marginal cost rather than short-run marginal cost, so prices below current marginal cost (and current average cost) do not necessarily imply that predatory behavior is occurring.(27) And newer theories suggest other motivations, such as those generated by irreversible investments, for profit-maximizing firms to price below average variable cost over some (perhaps relatively extended) periods.(28)

The pre-Uruguay Round antidumping statutes contained a legion of other serious biases, some of which were partially addressed by the Round (see section 3).(29) For example, individual transactions in the U.S. were compared to averages abroad and negative dumping margins were counted as zero, so any dispersion in the time series of prices produced a positive dumping finding, even if there was no dumping on a transaction-by-transaction basis.(30) In addition, the laws imposed minimum rates for profit (8%) and overhead (10%), regardless of the firm's actual profit or overhead. This tended to produce a higher constructed value, and thus a higher dumping margin, for firms with profit or overhead rates under the minimum. And misuse of the best-information-available process often meant that useful information submitted by foreign exporters was disregarded.(31) These provisions, as well as numerous others, resulted in positive dumping margins when no real dumping had occurred.

Table 1. U.S. Antidumping Cases by Industry, 1980-1989

Industry                         Number of Cases Filed

Chemicals                                 58
Food                                      16
Iron and Steel                           201
Machinery                                  8
Nonferrous Metals                         16
Textiles and Apparel                      15
Other                                    137

Total                                    451

Source: Boltuck and Litan (1991, Table 1-2, p. 3).

Given the focus on costs and the biases in the process, a dumping finding was almost a foregone conclusion, especially in the capital-intensive, highly cyclical industries that account for the majority of dumping cases (see Table 1). As Lester Thurow has noted, if "the law were applied to domestic firms, eighteen out of the top twenty firms in the Fortune 500 would have been found guilty of dumping in 1982" (Thurow 1985, p. 359). From this perspective, it is perhaps not surprising that a positive dumping margin was found in over 80% of all cases that reached a final determination in the early and mid-1980s.(32) So the dumping test often did not function as a binding constraint.(33)

Recent research has substantiated other distortions caused by the antidumping process, many of which are likely to continue in the post-Uruguay Round period. As one perhaps predictable example, empirical research has found that firms may not have to wait for a positive final determination in order to benefit from a dumping suit. Staiger and Wolak (1994b) report that "petitioning firms may enjoy import relief during the investigation period that amounts to about half of what they might expect from a positive final determination and duty imposition" (Staiger and Wolak 1994b, p. 101). There are several possible explanations for this result, including the retroactive calculation and imposition of duties at the end of the investigation (so that exporters and importers have an incentive to raise prices immediately once an investigation has begun) and the adverse effect of an investigation on competitive pressures within an industry. Whatever the explanation, the result has been that the antidumping laws are becoming increasingly attractive to firms seeking protection against imports - and not necessarily unfair imports.(34)

Because positive dumping margins are almost always found and because any significant fall in sales is sufficient for a positive injury finding (even if most of the fall in sales is caused by other factors), the antidumping laws are now effectively acting like market surge statutes, especially for moribund industries. Indeed, the laws are now sometimes referred to as the baby 201 laws.(35) This baby 201 tendency is the direct result of provisions within the import surge laws (like Presidential discretion, a higher injury standard, and the limited duration of remedies) that are intended to avert their misuse. In effect, then, the antidumping laws are functioning as an ersatz market surge system without the restrictions intended to prevent abuse.

According to many analysts, the antidumping laws are indeed being abused. This is perhaps most apparent in the filing of harassment cases intended to impose purely temporary trade restraints and legal costs on foreign exporters (Staiger and Wolak 1994a). Such cases are particularly effective because of the asymmetries in legal costs borne by domestic plaintiffs and foreign defendants. The current chairman of the ITC has been quoted as saying that he "can identify areas where there are abuses ... petitioners which use the trade laws purely as aggressive tools, rather than as a legitimate response to unfair trade practices."(36) Staiger and Wolak (1994b, p. 100) estimate that 10 of the 338 manufactured goods cases adjudicated between 1980 and 1985 were harassment or process cases - a small number, to be sure, but nonetheless worrisome.

In summary, the antidumping laws no longer have very much to do with the prevention of predatory pricing. Indeed, a recent OECD study examining U.S. dumping cases between 1979 and 1989 concluded that conditions were not even conducive to predatory pricing in over 85% of the cases that found dumping or that were terminated or suspended before a final determination.(37) It is perhaps worth stressing that the prevention of predatory pricing is a more restrictive standard than the prevention of price discrimination, which is itself a more narrow concept than preventing below-cost sales. And we do not even conduct our price-to-cost comparisons in an entirely disinterested manner and thus are at least three steps away from a purely predatory pricing rule.

As for the possible benefits of an antidumping law within a new trade theory environment, Boltuck and Litan (1991, p. 11) note that "there is no evidence that the antidumping law in the United States or in any other country is either designed or administered in a manner to ensure that its application is limited to such instances."

Countervailing Duty Laws

The first U.S. statute to address unfair international trade was an 1897 countervailing duty law addressing export subsidies (Committee on Ways and Means 1993, p. 53). This statute remained largely unchanged between 1922 (when foreign domestic subsidies were included) and 1979 (when the U.S. implemented the GATT Subsidies Code, which had been negotiated as part of the Tokyo Round).(38) The 1984 and 1988 trade acts made some other changes to the countervailing duty laws.

Under the countervailing duty law, the Department of Commerce determines whether imports into the U.S. are being subsidized, and the ITC determines whether material injury results from such subsidized imports. (The injury test is applied only in cases against GATT signatories.) As in antidumping, the procedure tends to produce biased margins, and the injury standard is low. Since it is relatively easy under these laws to show that a foreign firm has been subsidized in some way, the countervailing duty laws have become a popular protectionist sibling to the antidumping laws.

One contentious issue in countervailing duty law is the treatment of subsidies to firms that are subsequently privatized through auctions. It would seem to most economists that any subsidy granted prior to privatization would be capitalized into the privatization price. So the government would have been paid back, and the subsidy should not be countervailable, after privatization. Indeed, the Court of International Trade recently advanced such an argument.(39) This issue is of particular importance for the economies in transition and will be discussed further below.

Import Surge Laws

Section 201 is the primary U.S. market surge law, and it is generally consistent with the stringent requirements of the GATT safeguards clause (Article XIX). In particular, Section 201 requires that the ITC determine whether "an article is being imported into the United States in such increased quantities as to be a substantial cause of serious injury ... to the domestic industry..." (Committee on Ways and Means 1993, p. 452). If the ITC concludes that these requirements are satisfied, it then submits a suggested remedy to the President. The President has the discretion to approve the ITC's suggestion, to adopt a different remedy, or to do nothing at all. In making this decision, the President is instructed to evaluate whether a proposed action provides "greater economic and social benefits than costs." If a remedy is applied, it must be applied against imports from all trading partners (i.e., on a most-favored nation basis) and must be temporary (which is defined as anything lasting no more than eight years) (Committee on Ways and Means 1993, p. 464; Hansen and Prusa 1993, p. 38). Since it is much easier to obtain relief under the antidumping and countervailing duty statutes, Section 201 is now rarely used. No petitions, for example, were filed during the whole of 1993 (Office of the United States Trade Representative 1994, p. 105).

Section 337

Section 337 primarily addresses the infringement of patents, copyrights, trademarks, or mask works in the importation of goods into the United States (Committee on Ways and Means 1993, p. 444). Under this statute, the ITC investigates whether a violation has occurred and, if it reaches a positive determination, issues a remedial order to stop the imports, an order that the President can overrule. In 1989, a GATT panel concluded that parts of the Section 337 process violated the principle of national treatment because imported goods were subjected to a different legal system (distinct from the corresponding intellectual property rights statutes applying to domestically produced goods) solely on the grounds of their country of origin.(40)

Nonmarket Economies

Finally, the import trade statutes have a distinct set of procedures for so-called nonmarket economies (NMEs) like the countries of the former Soviet Union. The antidumping laws for NMEs are even more biased than those for market economies, and the NME market surge statute (Section 406) has a lower injury standard than Section 201.

NME dumping cases are determined using the surrogate country constructed cost measurement, in which factor costs from one or more surrogate market economies are combined with factor proportions from the NME to construct an average cost measure. The law stipulates that in choosing the surrogate country, Commerce should, "to the extent possible," rely on "market economy countries that are at a level of economic development comparable to that of the nonmarket economy country, and significant producers of comparable merchandise" (Committee on Ways and Means 1993, p. 389). Even a disinterested application of this methodology would be likely to produce dubious results. But the application of this statute has hardly been disinterested, and the results have often been farcical: Countries like the U.K. and Canada have been used as surrogates for China and Russia (Cass and Narkin 1991, p. 216). In a famous case involving Polish golf carts, Canada was chosen as a surrogate for Poland. But Canada had stopped producing golf carts before the investigation was completed. So after selecting Canada, the administering authority had to determine how much Polish golf carts would have cost in Canada if Canada had produced golf carts.(41)

Nonmarket economies also have unique status under the countervailing duty and market surge laws. They are currently exempt from countervailing duty cases because the courts ruled - in the Georgetown Steel case - that it was impossible to determine the level of subsidies in the distorted price environment that obtains in NMEs. But it is unclear whether subsidies granted while a country holds NME status are countervailable once the country graduates to market economy status. If they are, this would represent a substantial disincentive to moving to market economy status.

The market surge statute that applies exclusively to the NMEs, Section 406, is broadly similar to Section 201. But Section 406 has a lower injury standard (the surge in imports must be a "significant cause of material injury" rather than a "substantial cause of serious injury"), and remedies are applied only against imports from the NMEs, not against all countries (as under Section 201).

4. The Uruguay Round

Some of the most contentious battles of the Uruguay Round negotiations were waged over the unfair trade provisions. This section will briefly describe the relevant features of the Uruguay Round agreement.(42)

In antidumping, the Uruguay Round agreement generally moves world antidumping systems closer to U.S. procedures, though there are also some provisions that require changes in U.S. practice. Table 2 describes the more important changes in the U.S. laws. Examples of marginal improvements in the laws include the stipulations that profit and overhead calculations [TABULAR DATA FOR TABLE 2 OMITTED] must use actual data whenever possible (although some mischief is still possible when transaction are judged to be outside the ordinary course of trade, which is potentially a major loophole); that price comparisons must be on a transaction-to-transaction or average-to-average basis (although only in investigations, not reviews, and there are some exceptions even for investigations); that orders must be terminated after five years (unless such a revocation is "likely to lead to continuation or recurrence" of injury); and that evidence submitted by foreign exporters must be accepted and used (subject to some relatively weak requirements). While such changes represent a significant improvement in the law and do address some of the more egregious biases in the previous system, they nonetheless perpetuate the fundamental structure of the antidumping process, with its emphasis on preventing sales below average total cost.

The Uruguay Round agreement also addresses subsidies, which it divides into three categories: those that are prohibited (or so-called red light subsidies), those that are permitted unless they cause adverse trade effects (yellow light), and those that are permitted regardless of trade effects (green light). Red-light subsidies include export subsidies (both de jure and de facto) and subsidies contingent on local content requirements. Yellow-light subsidies are those not otherwise prohibited (under the red-light test) nor protected (under the green-light test) and are actionable only if they cause injury or serious prejudice to another country or somehow impair other trade provisions. Green-light subsidies include those not specific to particular enterprises or industries (that is, those that do not disproportionately benefit particular sectors), R&D subsidies (as long as the subsidies cover no more than 75% of the costs of industrial research or 50% of the costs of precompetitive development activity), regional development subsidies (as long as they are granted to a region considered to be disadvantaged on the basis of certain given objective criteria), and environmental subsidies (as long as they are nonrecurring, limited to 20% of environmental adaptation costs, and available to all relevant firms).(43) Subsidies found to contravene these rules can be disciplined by the new World Trade Organization (WTO) or countervailed by member countries.

Perhaps most importantly, the Uruguay Round agreement institutes a much stronger dispute settlement procedure under the auspices of the WTO. Under the new process, a panel decision will be adopted unless there is a unanimous agreement not to adopt it (under the previous GATT dispute settlement process, a panel decision was only adopted if there was a unanimous agreement to adopt it). An approved WTO panel decision against a particular country will require either a change in the relevant practice or the granting of compensation to other WTO members. (Compensation usually means that the other WTO nations are allowed to withdraw previous tariff concessions, providing one of the few types of compensation that can harm both the donor and the recipient.)

5. The Future: How Should We Improve Our Import Trade Laws?

The Uruguay Round brings our trade laws slightly closer to the principles adduced in section 2. But there is still an obvious gap. In policy interventions, as in statistical testing, there are both Type I errors (intervening when we shouldn't) and Type II errors (not intervening when we should). Our import trade laws seem to have an unnecessarily strong bias toward Type I errors and thus are inconsistent with President Clinton's often-quoted exhortation that "we must compete, not retreat." It is generally acknowledged that the laws need to be reformed. Even the chairman of the ITC admits that "we all know these laws can be improved."(44)

Because other countries are learning to mimic U.S. behavior, a more liberal set of import trade laws would bring indirect efficiency benefits - through less restrictive laws abroad - in addition to the direct effects (Congressional Budget Office 1994, p. 7). Indeed, U.S. exporters already face more antidumping cases than exporters from any other country [ILLUSTRATION FOR FIGURE 2 OMITTED],(45) and any biases in our laws that protect our import-competing firms are likely to redound to the detriment of our exporters.(46) And our most competitive exporters, the ones most likely to price below some given norm, are the ones most likely to be subjected to antidumping duties. Perpetuating unfair trade laws that are themselves unfair thus imposes substantial burdens on our consumers and on our most efficient exporters while protecting our least efficient import-competing firms.

Even if the import trade laws are not liberalized, there is still a variety of ways to improve them. Our antidumping and countervailing duty laws are now used for a variety of purposes for which they were not intended, while spurious accusations of unfairness obfuscate the real policy issues involved. Meanwhile, the domain of the laws is becoming ever more expansive. Trade attorneys, for example, sometimes argue that the antidumping laws must be used to combat foreign monopolies that restrict market access to U.S. firms, even though Section 301 includes an explicit provision to address such problems.(47) Basically, dumping has come to be defined as anything that can produce a positive dumping margin under the extant antidumping statutes. So even if changes are made only on a restrictiveness neutral basis, our system of trade laws could be more transparent and efficient if it functioned as it ostensibly should: with import surges handled under the import surge law, foreign anticompetitive practices handled under Section 301 (or in multilateral form), etc.

What specific reforms would improve the system? Numerous scholars of the laws have put forward suggestions that range from the incremental to the momentous.(48) Below I delineate one possible combination of these suggestions that would simplify the structure of the laws while moving us closer to the policy objectives advanced in section 2. As will become apparent, we do not necessarily abide by the restrictiveness neutral constraint.

Before turning to the specific proposals, I note three general points. First, a crucial issue in considering reform of the import trade laws is the relevance of the new trade theory effects mentioned in section 2. Within the economics community, there continue to be reservations, mainly based on political economy arguments, about the practical benefits of policies predicated on the new trade theory.(49) And any import trade law designed to capture the benefits suggested by the new trade theory seems prone to serious Type I errors: Numerous trade restrictions would undoubtedly be imposed when there are no rents to be captured or protected by the nation as a whole. My conclusion is that the precepts of the new trade theory are an important caveat and may be crucial in a limited number of specific cases, but attempts to exploit such effects within the import trade laws are likely to be both ineffective and costly. Second, any set of trade laws, regardless of how well the laws are formulated, is likely to induce wasteful rent-seeking behavior. Statutes offering even the possibility of protection inevitably engender rent-seeking activities that are both direct (e.g., lobbying) and indirect (e.g., manipulating output in order to make a positive injury finding more likely). Leidy (1994) describes the various ways in which such behavior manifests itself. We can only hope to minimize, not eliminate, such effects. Finally, although I do not explore the issue here, the inclusion of further improvements to our trade adjustment assistance programs could increase the political viability of any reform package while possibly yielding additional economic benefits.

Antidumping

The most extreme reform to the antidumping laws would be to merge them with our domestic predatory pricing laws. If predatory pricing is the primary policy concern raised by dumping (that is, if the weak form of predation is, in practice, unimportant), this suggestion seems to have substantial merit.

Many of the problems involved in designing a good antidumping law are also faced in designing a good domestic competition law, including the difficulties of measuring marginal cost (as opposed to average cost), the treatment of costs for new operations (startup costs), the nexus between market barriers and the probability of successful predation, and the definition of the relevant market. It seems foolish and inefficient to struggle with these questions twice, once in our domestic antitrust laws and then again in the international dumping laws. Furthermore, for reasons of fairness and simplicity, a single set of predatory pricing laws (for both domestic and international cases) seems preferable. Such a change in our laws would mirror the increasing harmonization and unification of competition policies internationally. In customs unions and free trade areas like the European Union and the Australia-New Zealand Free Trade Area, for example, antidumping laws no longer apply to trade between member countries. Instead, competition law standards are used (Council of Economic Advisers 1994, p. 240).

To be fair, there are some considerations that are specific to the international sphere. Additional complications in international cases include the problems of converting different currencies into a common one (i.e., choosing which exchange rates to use),(50) the rents accruing to foreign producers, the higher adjustment costs caused by the relative immobility of labor across national boundaries, and the possible involvement of sovereign foreign governments. But these difficulties are not intractable.(51)

Other, less severe, reforms to the antidumping laws are also possible. For example, it would not be difficult to remove or attenuate some of the more egregious procedural problems (for example, the focus on price-to-cost comparisons) simply by passing new laws or reinterpreting extant laws. Discretion could be incorporated into the process by allowing the President to veto dumping duties (although I note that the current quasi-judicial system was developed in part to protect the President from the political pressures inevitably associated with any decision on dumping duties). A more comprehensive definition of injury is an especially promising possibility. Finger (1993), for example, has suggested that the injury test should scrutinize the effect on the national economy rather than the petitioning industry. "Antidumping would thus become public policy rather than private policy" (Finger 1993, p. 70).

Countervailing Duty Laws

In principle, such laws can be beneficial if they are carefully crafted and implemented. But they must no longer be used as a form of protection against market surges. A stricter injury test and a higher standard of proof for establishing the existence of subsidies would be helpful in this regard. Francois, Palmeter, and Anspacher (1991) present a detailed set of the "more obvious reforms that are needed" in the countervailing duty procedures, including the elimination of hypocritical cases (that is, cases against practices also undertaken by the U.S. government), changes in the methodology used to calculate subsidies, and a variety of other specific recommendations.(52) The new World Trade Organization can play a useful role in promoting such reforms.

Import Surge Laws

Protection against market surges may be a legitimate feature of our trade law system, but such protection should probably not be conducted furtively through laws intended for other purposes. Most analysts agree that any surge protection should be transparent, undertaken only in the national economic interest, and betray no connotations of unfair trade. Section 201 seems largely to fulfill these requirements, although one critic has suggested some improvements (the title of his essay, "Throwing Deep: Trade Remedy Laws in a First-Best World," may betray the standard to which he is holding the surge law) (Trebilcock 1990, p. 249-250). Given the existing distortions in the import trade law system, the more urgent reforms surely involve the antidumping and countervailing duty laws.

Section 337

A separate set of laws to protect intellectual property rights violations during the importing process seems superfluous, at least to naive economists like ourselves. Can't all complaints about intellectual property rights be channeled through the corresponding domestic laws? In addition to parsimony and clarity, this would ensure that we are in full compliance with the 1989 GATT ruling.(53)

Nonmarket Economies

The nonmarket economy procedures are widely recognized as severely biased and difficult to defend, especially since the sui generis nature of the economies in transition makes comparisons with comparable market economies extraordinarily problematic. Both benevolence and hard-headed national interest, moreover, suggest that the reform process in the former Soviet Union and Eastern Europe should be supported. Open markets - "trade not aid" - seem the most auspicious and mutually beneficial means of providing such support. As Jeffrey Garten, the Undersecretary of Commerce for International Trade, has noted, "At present, most of the economies in transition are still considered non-market economies under the U.S. antidumping law. The United States believes that increased international trade is their best opportunity for development" (Garten 1994, p. 27-28).

One comprehensive reform to the nonmarket economy import trade laws was proposed by the Clinton administration in the spring of 1994. Under the proposal, which is still under consideration, trade complaints against the economies in transition, primarily the states of the former Soviet Union and Eastern Europe, would be channeled through a new import surge law. The antidumping, countervailing duty, and Section 406 processes would be temporarily suspended for these countries. The injury standard under this new law would be higher than in the antidumping code but lower than in Section 201. The ITC would evaluate injury claims and submit recommended remedies to the President. A positive injury finding would require some corrective action by the President (albeit not necessarily the one suggested by the ITC). This proposal seems to represent a substantial improvement over current practice.

6. Conclusion

Abuse of import trade laws represents one of the most ominous threats to a liberal international trading regime. There is a coherent argument for the existence of an extremely circumscribed set of such laws, but our statutes seem to have become divorced from such principles. The Uruguay Round represents a step in the right direction, but only a small step. Economists can play a useful role by continuing to point out the weaknesses in our import trade laws to policymakers and the public, while also proposing and debating possible reforms to improve the functioning of the system.

I thank John Anderson, Bob Cumby, Gary Horlick, Michael Meyer, Louise Keely, Caroline Levine, and Michael Orszag for helpful comments and/or data (subject to the usual disclaimer). I am especially grateful for the contributions of Peter Orszag, who developed many of these ideas with me during our time together at the Council of Economic Advisers and subsequently while he was completing his dissertation at the London School of Economics. The opinions expressed here do not necessarily represent those of the Council of Economic Advisers or any other agency of the U.S. government.

1 See Dale (1980, Appendix 3, p. 204) and Boltuck and Litan (1991, Table 1-1, p. 2). The discrepancy obtains despite the difference in time intervals (the first figure covers an 11-year period, whereas the second figure covers a 10-year period). Adjusting for the increased volume of merchandise imports during the later period attenuates, but does not eliminate, the increase. In the earlier period, 11.9 cases were investigated for every $100 billion (1987 dollars) of merchandise imports. In the later period, 13.2 cases were investigated for every $100 billion of merchandise imports. Thus, if unfair imports are assumed to be a constant proportion of total merchandise imports, there was still a significant increase in the use of antidumping procedures, even after accounting for the increased incidence of unfair trade.

2 Figure 1 depicts the total number of antidumping cases reported to the GATT Committee on Antidumping Practices minus the number reported by the U.S. Only signatories to the antidumping code within the GATT are required to file such reports, and many countries (including many GATT Contracting Parties) are not signatories. In addition, some signatories (e.g., the European Union) do not report some cases filed against nonsignatories. The figures presented thus represent a lower bound on the total number of antidumping cases filed outside the U.S. Sources: For 1980 through 1988, U.S. General Accounting Office (1990, p. 16). For 1989 through 1993, compilations based on GATT reports.

3 The unfair import trade laws are a subset of the import trade laws: The import surge law, for example, does not purport to address unfair trade.

4 In order to focus the discussion, we will concentrate on those laws aimed at restraining unfair import trade; laws that can result in import restraints but that are primarily aimed at removing impediments to exports - such as Section 301 will be discussed only tangentially. For a thorough examination of Section 301, see Bayard and Elliot (1994). We will also exclude laws dealing exclusively with agricultural concerns (e.g., Section 22 of the Agricultural Adjustment Act) or environmental issues (e.g., the Pelly Amendment).

5 There may be other justifiable policy concerns, like national defense objectives, but they lie outside the purview of this analysis.

6 For a relatively early analysis of dumping, see Viner's (1923) seminal work. He advocates the definition given in the text.

7 See, for example, Brander and Krugman (1983). In general, a domestic welfare gain from foreign dumping may be more likely if we are net importers of the product. In partial equilibrium, the net import assumption ensures that the consumer surplus gain from a lower price dominates the domestic producer surplus loss in the market for the imported good. But matters are more complicated once other markets are included in the analysis (assuming nonseparability of utility across commodities) and in a dynamic context.

8 Such concerns are only germane in an imperfectly competitive environment: Within a perfectly competitive world, antitrust issues are obviated because free entry obtains by assumption (ensuring zero profits in equilibrium).

9 For an overview, see, for example, Baker (1989).

10 This does not necessarily suggest that policymakers should be disproportionately concerned about international cases, however. The presence of a foreign producer in the market may indicate that the environment is less conducive to a successful predatory strategy, thus invalidating the ceteris paribus assumption above, if, for example, it indicates that entry by other foreign producers is relatively easy in the particular market (i.e., because import barriers are low). In other words, international predation may be both more deleterious and less likely.

11 Predation is an irreversible investment undertaken in an environment of uncertainty and thus has an option value that must be accounted for in the firm's predatory strategy. For an introduction to the options approach to irreversible investment under uncertainty, see Dixit and Pindyck (1994).

12 In this example, there are efficiency benefits to the predation process (the predator's marginal cost is below the prey's marginal cost, assuming the predator's marginal cost is no higher than her average cost). But these efficiency benefits must be compared to the distortions induced by the predator's increased market power in order to reach a conclusion as to whether the predation increases social welfare.

13 For basic introductions to the new trade theory, see Krugman (1986) and Helpman and Krugman (1989, Chapter 5).

14 In a Cournot model, for example, a reduction in one firm's marginal cost can result in an increase in its share of the market, even though the second firm does not vanish. So if the foreign firm's marginal cost is reduced because of new trade theory effects, it can capture a larger share of the rents in the market even if it does not drive the domestic firm out of business.

15 From a different but equivalent partial equilibrium perspective, our terms of trade improve. In general equilibrium, the impact on social welfare may be more complicated (e.g., if utility functions are not separable across commodities and if cross-elasticities are significant).

16 The laws may also discourage rent-seeking behavior from firms eager for the subsidies.

17 Available derivatives markets are not sufficient to hedge against large import surges.

18 The import surge laws are often referred to as escape clauses or safeguard statutes. But import surge laws seems to be the most accurate description of their substance.

19 For some reason, the antidumping and countervailing duty laws are not usually referred to by their section numbers.

20 In the discussion below, all references to predatory pricing and predation refer to the strong form, that is, to price reductions intended to drive other firms out of the market.

21 The law included a requirement that the dumping had to "be done with the intent of destroying or injuring an industry in the United States, or of preventing the establishment of an industry in the United States, or of restraining or monopolizing any part of trade and commerce in such articles in the United States" (Committee on Ways and Means 1993, p. 417).

22 The legislative history of the 1921 Act suggests that, despite the formal language of the law, the prevention of predatory pricing was still its primary objective. As the committee report states, the Act "protects our industries and labor against a now common species of commercial warfare of dumping goods ... until our industries are destroyed, whereupon the dumping ceases and prices are raised at above former levels to recoup dumping losses" (quoted in Dale 1980, p. 12). See also Nivola (1993, p, 31).

23 Congressional hearings held in 1968 to consider the Kennedy Round antidumping codes clearly indicate that protagonists then viewed the prevention of price discrimination as the primary objective of the antidumping laws. In his opening statement, for example, Senator Russell Long is quoted as saying, "The Antidumping Act ... is designed to protect against the rankest kind of commercial injustice - price discrimination by foreign competitors ..." (emphasis added) (Committee on Finance 1968, p. 1). See also Cass and Narkin (1991, pp. 208-212).

24 The methodology used in cases involving nonmarket economies is different.

25 The Commerce Department's pre-Uruguay Round methodology disregarded all foreign price data and used the constructed value index if more than 90% of foreign sales were deemed to be below cost.

26 The Areeda-Turner predatory pricing criterion recognizes that sales below average total cost are not necessarily predatory (see Areeda and Turner 1975, pp. 697-733).

27 See Spence (1981). Spence's model assumes no uncertainty and a zero discount rate. For a discussion of learning curve effects with uncertainty and a positive discount rate, see Dixit and Pindyck (1994, pp. 339-353).

28 See Dixit and Pindyck (1994, pp. 292-293).

29 For more detail on many of the biases in the pre-Uruguay Round dumping laws, see the collection of essays in Boltuck and Litan (1991). The collection also includes an essay by an enthusiastic supporter of the statutes (see Stewart 1991, pp. 288-330).

30 Suppose that sales in both the foreign market and the U.S. market occurred on two dates. On the first date, the good was sold for $10 in the U.S. market and for the equivalent of $10 in the foreign market. On the second date, the good was sold for $5 in the U.S. market and for the equivalent of $5 in the foreign market. Clearly, there was no dumping; the goods were sold for the same price on both occasions. But the transaction-to-average comparison process would nonetheless have produced a positive dumping margin: The average price for foreign sales was $7.50, so the second sale in the United States was dumped, with a margin of $7.50 - $5.00 = $2.50. The first sale in the U.S. had a dumping margin of -$2.50, so that if the two individual margins were averaged, the correct answer (zero) would be obtained. But instead the first dumping margin would have been set to zero, and the final average dumping margin would therefore have been $1.25 (=$2.50/2).

31 A recent article in The Washington Post noted that "the administrative burden simply of furnishing the required information within the required time in the required form to the Department of Commerce has become so overwhelmingly difficult that more and more foreign companies are either unable or unwilling to try ... [C]ompanies of the stature of Matsushita, SKF, and Toshiba find compliance with the Commerce Department's demands impossible ..." (Thompson 1994). Not surprisingly, there is some evidence that cases that did not use information from foreign exporters resulted in larger dumping margins than cases that did. Baldwin and Moore (1991, Tables 701, 703) find that the average dumping margin for final determinations using information from foreign firms was 27.9%, whereas the average dumping margin for final determinations that did not use such information was 66.7%. In a pooled regression of all final dumping margins, the dummy variable for the lack of foreign information was positive (0.383), with a t-statistic of 4.19. But it is unclear whether these results are causal or merely an example of selection bias: It could be that the foreign firms most unfamiliar with the antidumping process or with the least incentive to comply with that process were also the ones with the largest dumping margins.

32 Murray (1991, Table 2-3, p. 46). The figure applies to cases initiated after 1979 and concluded before 1986. It excludes cases that were terminated or suspended. According to a press report, positive determinations have been even more likely recently: An article in Forbes magazine alleges that dumping was found in 97% of the cases decided between 1988 and 1992 (see Frum 1992).

33 We should note, however, a possible selection bias: The cases in which dumping would not be found may not be observed as frequently because petitioners can predict the outcome and therefore a disproportionate number of such cases are not filed. This criticism does not seem relevant to Lester Thurow's observation.

34 Between 1963 and 1979, U.S. firms filed 532 antidumping and countervailing duty cases and 75 import surge cases. Between 1980 and 1988, they filed over 700 antidumping and countervailing duty cases and just 19 import surge cases (Hansen and Prusa 1993, p. 35).

35 The sobriquet applies to the countervailing duty laws also (Boltuck and Litan 1991, p. 11).

36 See Financial Times (10 August 1994).

37 See OECD Working Party No. 1 (1994).

38 The most important reform in 1979 was the introduction of a mandatory injury test for all signatories to the Code.

39 The case involved the U.K.'s United Engineering Steels, a joint venture between British Steel and Guest, Keen, and Nettlefolds. The Court overturned Commerce's decision to countervail subsidies granted prior to privatization (see Financial Times 15 June 1994). This issue is of crucial importance in assessing the benefits of the transfer of nonmarket economies to market economy status. See discussion in the text that follows.

40 See Office of the United States Trade Representative (1994, p. 104) and Jackson (1989, p. 192).

41 Technically, this case was adjudicated using a surrogate country price measure: Canada's golf cart price, rather than factor costs, was used to construct the foreign market value. But since Canada did not produce golf carts, the price had to be constructed. In effect, then, the only difference between the surrogate country price measure and the surrogate country constructed cost measure in this case was that Canada's factor proportions were used instead of Poland's. In an administrative review, the surrogate country was switched to Spain.

42 For an overview of the results of the Uruguay Round and an introduction to the operations of the World Trade Organization, see Jackson (1994).

43 The traffic light rules are somewhat different for less developed countries.

44 See Financial Times (10 August 1994).

45 Figure 2 depicts the number of antidumping cases filed against exporters from the listed countries and reported to the GATT Committee on Antidumping Practices between January 1989 and June 1993. Only signatories to the antidumping code within the GATT are required to file such reports, and many countries (including many GATT Contracting Parties) are not signatories. In addition, some signatories (e.g., the European Union) do not report some cases filed against nonsignatories. The figures presented thus represent a lower bound on the total number of antidumping cases against the given countries. It is possible that the rankings for the total number of antidumping cases, filed by GATT and non-GATT signatories, would differ from the rankings given here. Source: Compilations based on GATT reports.

46 The biases in our own laws vitiate our position in criticizing similar biases in other nations' laws. It seems highly unlikely that the U.S. government would file complaints with the WTO against other countries for effectively mimicking our laws.

47 See Jackson (1989, p. 212). Section 301 includes in its definition of "unreasonable acts" the "toleration by a foreign government of systematic anticompetitive activities by private finns or among private firms in the foreign country that have the effect of restricting ... access of United States goods ...."

48 See, for example, Jackson (1989, pp. 242-244), Boltuck and Litan (1991), Trebilcock (1990), Hart (1990), and especially Finger (1993, Chapter 4).

49 See, for example, Krugman (1987, 1994). Francois (1992, p. 185) concludes that, even in the presence of increasing returns, "on net optimal commercial policy may still best be approximated by free trade."

50 This problem can actually arise in purely domestic antitrust cases, albeit in less severe form, if inputs are imported.

51 We should note that merging the antidumping laws with the domestic competition laws would probably reduce rent-seeking activities but would not eliminate them: The antitrust laws have also induced much rent-seeking behavior. William Baumol (1993, pp. 78-82) provides an interesting set of examples, including the London Fuller's case of 1298, on the national economy rather than the petitioning industry. "Antidumping would thus become public policy rather than private policy" (Finger 1993, p. 70).

52 See Francois, Palmeter, and Anspacher (1991, pp. 130-133).

53 According to officials in the Office of the U.S. Trade Representative, the Uruguay Round implementing legislation brings Section 337 into compliance with the GATT ruling. But it is not clear that a WTO panel would concur.

References

Areeda, P., and D. Turner. 1975. Predatory pricing and related practices under Section 2 of the Sherman Act. Harvard Law Review 88:697-733.

Baker, J. 1989. Recent developments in economics that challenge Chicago school views. Antitrust Law Journal 58:645-55.

Baldwin, R., and M. Moore. 1991. Political aspects of the administration of the trade remedy laws. In Down in the dumps: Administration of the U.S. unfair trade laws, edited by R. Boltuck and R. Litan. Washington, DC: Brookings Institution Press, pp. 253-87.

Baumol, W. 1993. Entrepreneurship, management, and the structure of payoffs. Cambridge, MA: MIT Press.

Bayard, T., and K. Elliot. 1994. Reciprocity and retaliation in U.S. trade policy. Washington, DC: Institute for International Economics.

Boltuck, R., and Litan, R. 1991. Down in the dumps: Administration of the U.S. unfair trade laws. Washington: Brookings Institution Press.

Brander, J., and P. Krugman. 1983. A 'reciprocal dumping' model of international trade. Journal of International Economics 15:313-21.

Brander, J., and B. Spencer. 1984. Tariff protection and imperfect competition. In Monopolistic competition and international trade, edited by H. Kierzkowski. Oxford: Clarendon Press, pp. 194-206.

Brander, J., and B. Spencer. 1985. Export subsidies and market share rivalry. Journal of International Economics 18: 83-100.

Cass, R., and S. Narkin. 1991. Antidumping and countervailing duty law: The United States and the GATE In Down in the dumps: Administration of the U.S. unfair trade laws, edited by R. Boltuck and R. Litan. Washington, DC: Brookings Institution Press, pp. 208-12.

Committee on Finance. 1968. U.S. Senate international antidumping code. Ninetieth Congress, Second Session, June 27.

Committee on Ways and Means. 1993. Overview and compilation of U.S. trade statutes. Washington, DC: Government Printing Office.

Congressional Budget Office. 1994. A review of U.S. antidumping and countervailing duty law and policy. Washington, DC: Congressional Budget Office.

Council of Economic Advisers. 1994. Economic report of the President 1994. Washington, DC: Government Printing Office.

Dale, R. 1980. Anti-dumping law in a liberal trade order. New York: St. Martin's Press.

Dixit, A. 1987. How should the United States respond to other countries' trade policies? In U.S. trade policies in a changing world economy, edited by R. Stem. Cambridge, MA: MIT Press, pp. 245-82.

Dixit, A., and R. Pindyck. 1994. Investment under uncertainty. Princeton, NJ: Princeton University Press.

Financial Times, 14 June 1994. Steel duties rejected. p. 4.

Financial Times, 10 August 1994. US trade sleuth in search of fairness. p. 4.

Finger, J. M. 1993. Antidumping: How it works and who gets hurt. Ann Arbor: University of Michigan Press.

Francois, J. 1992. Optimal commercial policy with international returns to scale. Canadian Journal of Economics 25: 184-95.

Francois, J., N. D. Palmeter, and J. Anspacher. 1991. Conceptual and procedural biases in the administration of the countervailing duty law. In Down in the dumps: Administration of the U.S. unfair trade laws, edited by R. Boltuck and R. Litan. Washington, DC: Brookings Institution Press.

Frum, David. 1992. Dump it. Forbes, 28 September, p. 64.

Garten, J. 1994. The United States in a changing world economy: Recent speeches. Washington, DC: U.S. Department of Commerce.

Hansen, W., and T. Prusa. 1993. Does administrative protection protect? A reexamination of the U.S. Title VII and escape clause statutes. Cato Review of Business and Government Regulation Winter:35-6.

Hart, M. 1990. Idealism versus pragmatism: Policy and the academic analyst. In Fair exchange: Reforming the trade remedy laws, edited by M. Trebilcock and R. York. Toronto: CD Howe Institute.

Helpman, E., and P. Krugman. 1989. Trade policy and market structure. Cambridge, MA: MIT Press.

Jackson, J. 1989. The world trading system: Law and policy of international economic relations. Cambridge, MA: MIT Press.

Jackson, J. 1994. Managing the trading system: The World Trade Organization and the post-Uruguay Round GATt agenda. In Managing the world economy: Fifty years after Bretton Woods, edited by R Kenen. Washington, DC: Institute for International Economics.

Krugman, P. 1984. Import protection as export promotion: International competition in the presence of oligopoly and economies of scale. In Monompolistic competition and international trade, edited by H. Kierzkowski. Oxford: Clarendon Press. Krugman, R 1986. Strategic trade policy and the new international economics. Cambridge, MA: MIT Press. Krugman, R 1987. Is free trade passe? Journal of Economic Perspectives 1:131-44.

Krugman, P. 1994. Empirical evidence of the new trade theories: The current state of play. In New trade theories: A look at the empirical evidence. Centre for Economic Policy Research.

Leidy, M. 1994. Trade policy and indirect rent seeking: A synthesis of recent work. Economics and Politics 6:97-115.

Murray, T 1991. The administration of the antidumping duty law by the Department of Commerce. In Down in the dumps: Administration of the U.S. unfair trade laws, edited by R. Boltuck and R. Litan. Washington, DC: Brookings Institution Press.

Nivola, P. 1993. Regulating unfair trade. Washington, DC: Brookings Institution Press.

OECD Working Party No. 1. 1994. Chapter 2: Census and analysis of antidumping actions in the United States. In Working Party No. 1 on competition and international trade: Antidumping and competition policy, draft (6/4/ 94). Washington, DC: OECD.

Office of the United States Trade Representative. 1994. Trade policy agenda and 1993 annual report of the President of the United States on the Trade Agreements Program. Washington, DC: U.S. Trade Representative's Office.

Spence, M. 1981. The learning curve and competition. Bell Journal of Economics 12:49-70.

Staiger, R., and F. Wolak. 1994a. The determinants and impacts of antidumping suit petitions in the United States: An industry level analysis. In New trade theories: A look at the empirical evidence. Centre for Economic Policy Research.

Staiger, R., and F. Wolak. 1994b Measuring industry-specific protection: Antidumping in the United States. In Brookings papers: Microeconomics 1994. Washington, DC: Brookings Institution Press, pp. 51-118.

Stewart, T. 1991. Administration of the antidumping laws: A different perspective. In Down in the dumps: Administration of the U.S. unfair trade laws, edited by R. Boltuck and R. Litan. Washington, DC: Brookings Institution Press, pp. 288-330.

Thompson, T 1994. Dumping on trade. The Washington Post, 18 July, p. A19.

Thurow, L. 1985. The zero sum solution: Building a world-class American economy. New York: Simon and Schuster.

Trebilcock, M. 1990. Throwing deep: Trade remedy laws in a first-best world. In Fair exchange: Reforming the trade remedy laws, edited by M. Trebilcock and R. York. Toronto: CD Howe Institute, pp. 249-50.

U.S. General Accounting Office. 1990. Internal trade: The use of the GATT Antidumping Code. Washington, DC: U.S. General Accounting Office.

Viner, J. 1923. Dumping: A problem in international trade. Chicago: University of Chicago Press.

Joseph E. Stiglitz is Senior Vice President and Chief Economist, World Bank, 1818 H Street, NW, Washington, DC 20433, USA; e-mail [e-mail].