Dumping: Still a Problem in International Trade
Thomas R. Howell
Dumping is the export of products at less than "normal value," often defined as the price at which those products are sold in the home market. Since its inception, the General Agreement on Tariffs and Trade (GATT) has authorized signatories to apply duties to offset dumping when it causes, or threatens to cause, material injury to an industry in the territory of a GATT member.1 National antidumping legislation dates from well before the GATT. For example, the United States passed its first antidumping statutes in 1921.
Despite their longevity, antidumping measures are frequently subject to sharp criticism, especially from academic economists. Indeed, some observers advocate their complete elimination, raising the question whether dumping itself is a problem sufficiently serious to warrant retention of the antidumping regime provided for under the GATT. This paper notes that antidumping measures, like any complex regulatory regime, may give rise to anomalous or undesirable results in some cases, but argues that dumping itself remains a "problem in international trade," as described by Jacob Viner in his seminal 1923 study of the subject. As such, dumping requires continued regulation, especially for countries with relatively open national markets.
The existence of price discrimination between domestic and export markets generally indicates the presence of a market distortion in the home market, such as import barriers, a monopoly or cartel, or some combination of these factors that gives domestic producers the ability to maintain domestic prices at a level higher than export prices. Under such circumstances, dumping is a mechanism through which competitive outcomes are determined, in effect by the distortion itself, not the relative competitiveness of individual producers. In the short run, dumping enables protected firms to run their facilities at higher utilization rates than would be economically feasible in an open market, giving them a major cost advantage unrelated to their comparative cost competitiveness. Over the long run, dumping can deter investment in the market where it is occurring and, conversely, may well foster increased investment in the protected market. Over time, through such dynamics, dumping may permit an initially less efficient (but protected and cartelized) industry to displace an equally or efficient competitor, that is, not benefiting from a protected home market.
Because dumping can result in the erosion or destruction of national industries for reasons unrelated to normal market competition, simply permitting dumping to occur without any regulation could endanger the political consensus which supports the current liberal multilateral trading system. Friction arising out of dumping can become particularly acute when dumping injures or destroys industries regarded as vital to national economic well-being and national security, a phenomenon which has been observable at a number of points in this century.
Fundamentally, the controversy surrounding antidumping is a symptom of a larger phenomenon, the divergence which exists between various national markets with respect to competition policy and which has frustrated all attempts at consensus for at least half a century. Antidumping measures have been assigned, more or less by default, the task of addressing specific problems created by this divergence. They are admittedly an imperfect tool. But until broader national differences with respect to competition policy are reconciled, these measures remain essential to the world trading system, acting, in the words of John Jackson, as an "interface mechanism... necessary to allow different trade systems to trade harmoniously."
* * *
Today's open multilateral trading system stands as one of the greatest achievements of the generation of statesmen that laid the foundations of the postwar world order. The legal underpinning of this system is provided by the General Agreement on Tariffs and Trade (GATT) and its ancillary agreements and codes, currently administered by the newly formed World Trade Organization. The GATT has made possible the progressive liberalization of world trade through the basic mechanism of binding commitments by signatories to reduce trade barriers on a most-favored-nation basis. The GATT has survived, however, in significant part, because its framers were wise enough to recognize that the system would not be sustainable in the absence of certain exceptions to the general com-
mitments undertaken by the signatories. These exceptions, which include "escape clause" provisions, special rules for developing countries, and antidumping and countervailing duty measures, have functioned as interface mechanisms to soften the dislocations that have occurred as the reduction in border restrictions has brought differing national economic systems into progressively closer competitive contact. Without the existence of these mechanisms, given
the politically sensitive subject of international trade . . . the General Agreement might never have been concluded or might never have endured in the face of the pressures that have buffeted it.2
One of the most significant exceptions to the basic GATT principle for most favored nation treatment authorizes contracting parties to apply duties "in order to offset or prevent dumping." GATT Article VI provides that
dumping, by which products of one country are introduced into the commerce of another country at less than the normal value of the products, is to be condemned if it causes or threatens material injury to an established industry in the territory of a contracting party or materially retards the establishment of a domestic industry.3
The right to apply antidumping measures was an important element in the original consensus that made the formation of the GATT possible, and the contracting parties to the GATT subsequently elaborated a complex system of rules and procedures pursuant to which members may apply antidumping duties in appropriate cases.4 Within these parameters, most major trading nations have enacted antidumping rules. It is often overlooked that the most active users of antidumping measures have been GATT members with the more open markets—countries such as Australia, Canada, the European Union, and the United States. As a number of newly industrializing nations liberalize their trade regimes, they are becoming more active in applying antidumping measures.
Antidumping policy is now the subject of scathing attacks from many quarters, including prominent figures in law, business, and academia. 5Forbes characterizes the antidumping laws as tools that U.S. firms use to push foreign firms into "a quick descent into legal hell" as they "lustfully anticipat[e] a price hike" for domestic consumers. 6 Claude Barfield of the American Enterprise Institute calls antidumping measures ''the chemical weapons of the trade wars," a "system
of price-fixing cartelization . . . that stacks the deck in favor of local producers against their foreign competitors."7 In a 1993 Wall Street Journal commentary, James Bovard branded the U.S. antidumping laws as a "fraud," castigating the laws for their "hypocrisy and absurdity," which enable "a few greedy producers" to invoke remedies that cause the U.S. government ''to inflict unlimited amounts of unfairness in the name of fair trade."8 Although academic critics are usually less colorful in their choice of words, the sheer number of eminent economists who have attacked antidumping is impressive. One might ask why laws so odious have not been quickly repealed. This has not occurred in the United States at least, according to some critics, because of "lobbying" by "protected U.S. producers," 9 because of "bureaucrats" at the Department of Commerce seeking "to flaunt the fact that there are few restraints on [their] power over foreign companies,"10 and, perhaps inevitably, because of the baneful influence of "lawyers."11
Curiously, there has been little response to the rising chorus of criticism of antidumping policy. Apart from a few obscure monographs and articles, little has been published defending the rationale for antidumping policy since Professor Jacob Viner, one of the draftsmen of the original U.S. antidumping laws, produced what remains the seminal work on the subject in 1923.12
The purpose of antidumping measures is to offset economic injury caused by the commercial practice of dumping. Although antidumping measures can be, and sometimes are, applied in an arbitrary, irrational, or unnecessarily burdensome manner, the same can be said of any major regulatory program or system of legal redress, and such problems do not, by themselves, constitute a basis for
scrapping an entire system. The real issue is whether dumping itself is a practice that warrants continuing restriction by national governments. The common strand that unites most critiques of antidumping is the extent to which they avoid that question, tending to minimize or dismiss altogether the phenomenon of dumping itself as not warranting serious examination. In the thousands of pages that have been written attacking antidumping, it is a challenge to find any detailed case study of an actual episode of dumping or an examination of its problematic aspects and implications. Were such inquiries more common, it would be evident that dumping remains "a problem in international trade" that warrants the continued existence of workable regulatory constraints on the practice.
The term "dumping" has enjoyed a casual business use for at least two centuries and is still loosely applied in a lay context to a variety of export practices involving low pricing. Jacob Viner' s groundbreaking 1923 work proposed a precise definition, "price discrimination between national markets," that has gained general acceptance as the definition of "classic" dumping and is now embodied in the GATT and national antidumping legislation. Under classic dumping, a seller charges higher prices in the home market than in export markets, or, much less commonly, charges higher prices in one export market than in another. The dumper is able to maintain a price differential because some factor or combination of factors separates the two markets—generally either the sheer distance between the markets or a protective barrier around the market where the higher price is charged, coupled with restraints on competition in the latter market.13 The first antidumping statutes, which were enacted between the end of the nineteenth century and the early 1920s, were directed against classic dumping only, but during the postwar era, their scope has been expanded, in effect, to embrace some types of export sales that are made below the cost of production, notwithstanding the absence of price discrimination between national markets.
THE EFFECTS OF DUMPING
Dumping leads to the erosion and in some cases the disappearance of industries in markets where dumping is occurring for reasons unrelated to the relative competitiveness of those industries—put most simply, dumping enables less efficient firms to prevail over more efficient firms in international competition. Competitive outcomes are determined by market distortions, that is, the factors that make dumping possible, rather than the relative competitiveness of individual producers. This occurs for two reasons:
Capacity utilization. Over the short run, other things being equal, dumping firms tend to enjoy lower unit costs than comparable firms in markets where dumping is occurring because dumpers can operate their plants at higher rates of capacity utilization—a factor that often has a far greater impact on cost than any other consideration. Firms in the market where dumping is occurring cannot respond in kind if the market of the dumper is closed to them. In this way, a relatively inefficient plant run at 100 percent utilization rates may well enjoy lower unit costs than a state-of-the-art facility run at a 50 percent rate.
Investment deterrent. Over the longer term, dumping discourages investment in markets where dumping is occurring, and, at the same time, encourages higher levels of investment in the protected markets from which dumping is taking place. This occurs because investment risks are higher, and returns lower, in markets where dumping is taking place, and risks are lower, and returns higher, in the protected market from which dumping is taking place. The short-run cost advantage that dumping firms enjoy is thus eventually translated into a capital and technological advantage as investment dries up in the one market and intensifies in the other.
The fact that unconstrained dumping can gradually lead to a shift in competitive advantage has implications that extend beyond the firms directly affected. A given nation's economic well-being, standard of living, and security are all determined in significant part by the composition of its industrial base. The ultimate implication of the competitive dynamics of dumping is that the industrial base can be altered in deleterious ways as a result of market distortions abroad, such as protected markets and cartels, that make dumping feasible. Because such distortions can be deliberately created and manipulated, whether by governments or by private syndicates enjoying the toleration or tacit encouragement of state authorities, the decision to permit unrestricted dumping is a decision to allow a national economy to be shaped by anticompetitive strategies and market distortions that are engineered in other countries. Although experience has shown that GATT signatories will accept, as part of the price of an open trading system, the need for adjustment by domestic industries that have lost international competitiveness, it is quite another matter to expect signatories to accept the burdens of adjustment that arise out of anticompetitive practices in other countries. It is unlikely that many nations would accept such a result for any sustained period. It is equally unlikely that a political consensus could be sustained for any multilateral regime that attempted to enforce it through proscriptions on national antidumping measures.
DUMPING UNRESTRICTED: THE BRITISH CASE
Would the world trading system as a whole, and its members individually, be better off if all antidumping measures were eliminated and dumping allowed to
occur without interference? A distinguished contemporary critic of antidumping policy, J. Michael Finger of The World Bank, argues that
The most appealing option is to get rid of antidumping laws and to put nothing in their place. Then all of the evils of such policy—its power politics, its bad economics, and its corrupted law—would be eliminated.14
Is Finger's proposal a sound one? Fortunately, this question is not altogether speculative, since trade between industrialized nations did occur for at least half a century before the widespread adoption of antidumping measures in the 1920s. Dumping was pervasive and its dynamics and effects widely reported and discussed. While many countries (including the United States) were relatively unaffected by dumping because high tariff walls severely limited import competition, Britain offers an example of a major, fully industrialized country that elected to avoid any policy action against dumping and to remain, in effect, an open "dumping ground" for a protracted period. Britain's rationale for adhering to free trade in the face of widespread dumping in her domestic and overseas markets was based on many of the lines of reasoning that are used today by those who urge the complete elimination of antidumping measures. Britain's disheartening industrial and commercial performance during this period, which saw the precipitous competitive decline of the industries most severely affected by dumping and a disastrous (and very near fatal) erosion of the country's strategic industrial base, is now an established historical fact. Although the "British disease" was the product of an extraordinarily complex tangle of economic and social problems of which dumping comprised only one strand, the historical record yields enough evidence of the harmful effect of unrestricted dumping on British industry to cast serious doubt on the wisdom of the policy that was followed.
In 1870 Britain accounted for more of the world's manufacturing output than any other nation, its industries boasted the lowest costs and the most advanced production technologies, and its banks and shipping firms dominated world commerce. Britain's prosperity and commercial dominance appeared to validate the philosophy of the Free Trade movement that, after a series of intense political battles, had in the 1840s succeeded in clearing away most of Britain's import restrictions. 15 In retrospect, however, it is evident that Britain's success prior to 1870 was attributable, in substantial part, simply to the fact that the country had industrialized before any other nation. Beginning in the 1870s, and growing in intensity thereafter, the rapidly growing manufacturing industries of Germany and the United States mounted a commercial assault on traditional British mar-
kets. In contrast to Britain, both the United States and Germany were avowedly protectionist; by 1880 both national markets were surrounded by high tariff walls. In addition, in both of these countries, highly organized and sophisticated anticompetitive industrial combinations were formed for the purpose of reducing competition and exploiting their partial or complete monopoly power. In the United States, so-called "trusts" regulated output and prices in many major manufacturing industries, and in Germany, manufacturing was dominated by kartells (cartels) in which price and output restrictions were maintained through legally enforceable contractual commitments.16 It was the standard practice of both the American trusts and the German kartells to engage in large-scale dumping as a deliberate export strategy.17
British industries did not exist in a competitive milieu that permitted them to respond in kind to this challenge. They could do nothing to reopen the American or German markets that had been lost to them, and they lacked both the protected home market and the organized character needed to engage in dumping on an American or German scale. 18 American and German firms not only captured sales from British firms, but began surpassing British industry in the level of industrial technology, productivity, and economies of scale. British producers confronted a strategic dilemma for which the Free Trade doctrine offered no obvious answers. Britain's Tariff Commission summarized this quandary in 1904 as follows:
[I]t is the control of the home market which their tariffs give to foreign countries, combined with the facilities for exportation which they secure through their trusts and kartells, and the free access to the British market, which is the condition of their rapid progress relative to the United Kingdom. These tariffs were, in many instances, deliberately adopted to shut out British products which came into competition with home manufacturers. Their adoption has been followed by (i.) the extinction or diminution of British competition in the foreign protected markets; (ii.) the closing of British works or of departments of British works which depended on these markets; (iii.) the rapid growth of the foreign
competing industry; (iv.) the appearance in the British market of the products of that industry at prices which the British manufacturer cannot touch. Thus, the positions of the United Kingdom and its most powerful competitors have been reversed.19
Britain's eroding competitive position relative to two dynamic protectionist powers began to foster dissent from the prevailing free trade orthodoxy, and in 1895, the issue was moved to the center of the nation's political arena by the governing Conservative Unionist party.20 In that year, Joseph Chamberlain, the government's Colonial Secretary and an avowed imperialist, began a crusade against free trade in favor of an imperial customs union that would establish a wall of protective tariffs around Britain and the Empire. Conservative Unionist Prime Ministers Lord Salisbury (1895-1902) and Arthur Balfour (1902-1906) shared Chamberlain's skepticism about free trade and were concerned over mounting evidence of Britain's economic decline relative to Germany and America, but were ultimately unwilling to commit their party and their country to a renunciation of free trade. Instead, Balfour sought a middle ground, the selective imposition of retaliatory tariffs against trading partners that practiced restrictive trade which hurt British industry.
The British debate over dumping at the turn of the century closely parallels the current controversy in the United States at the century's end. The Chamberlain and Balfour factions singled out "dumping" by foreign "trust system[s] working behind tariffs"21 and argued that dumping was injuring or destroying key industries on which Britain's economy and security rested.22 Dumping, it was argued, placed domestic industries at a cost disadvantage, eroded producers' profits, and jeopardized "the provision of adequate capital for carrying on great modern industries."23 Nonsense, responded the Free Traders. British industry was still faring well under free trade.24 The industries complaining of dumping were
seeking to blame imports for problems that were really of their own making. Dumping was actually a positive good, not only because it provided a stimulus to such firms to reform their ways, but because it provided cheap inputs for many other industries, lowering their costs.25 The alleged threat to "staple" industries was brushed off as exaggerated; moreover, it was pointed out, the disappearance of staple industries was more than offset by the appearance of new industries utilizing dumped inputs.
The arguments against antidumping measures carried the day in turn-of-thecentury Britain, and the Conservative Unionist assault on free trade served only to bring an electoral debacle upon the governing party.26 Britain took no measures to restrict dumping until a number of years after World War I. But while the Free Traders won the political debate, what were the consequences? Did Britain ultimately fare better or worse for having allowed itself to exist as a "dumping ground" until well after World War I?
Dumping and the Erosion of British Competitiveness
From the perspective of the late twentieth century it is evident that in the period 1880-1914 British industry was moving on a path of decline relative to the industries of the United States and Germany, a trend that would become increas-
ingly obvious as the twentieth century progressed.27 Britain's decline from the zenith of the mid-1800s has been extensively examined, but its causes remain something of an enigma.28 Dumping in British markets by foreign cartels was not the sole or even the primary proximate cause of Britain's relative industrial decline, but it does not follow that dumping played no role, or that Britain was, on the whole, better off for having permitted unrestricted dumping. Dumping was identified by many contemporary partisans in the trade debate as a significant factor contributing to both the erosion of British cost competitiveness and the inadequate levels of British capital investment.29 Both of these factors have been cited by subsequent generations of scholars as important, if not central, elements underlying British industrial decline.30
The British iron and steel industry was the centerpiece of the British debate over dumping, and its particular experience with dumping in this industry is probably of greatest relevance to the current dumping controversy because dumping in this industry was more pervasive in its extent and effects than in most other sectors, and because of steel's central importance to Britain's economy and national defense. The slump in Britain's position as a steel producer in the 1890s "was particularly alarming,"31 given steel's status at the time as the most important of all strategic industries, and it was addressed and analyzed by virtually all of the partisans on both sides of the trade controversy. While Free Traders argued that there was insufficient evidence that dumping was substantially injuring domestic producers, 32 the weight of evidence from the period makes it clear that by the mid-1890s, British steelmakers were under attack from low-priced Ger-
man and American steel across a broad product range in both overseas and domestic markets, and that they were being badly hurt by the loss of business.33
The mere loss of sales described in contemporary reports does not by itself necessarily indicate the existence of a problem that would have warranted a change in government trade policy. The American-German onslaught might, for example, simply have reflected the emergence of more efficient competitors abroad, confronting British producers with the choice of adapting to remain competitive or getting out of the business, a point that was in fact made many times during the debate over dumping. However, a close examination of the situation confronting British steelmakers at the turn of the century suggests that there was more at work than simply shifting comparative advantage. The combination of high tariffs, cartels, and the incentive to sell products below average cost had powerful effects both on immediate commercial positions and long-term relative competitiveness. Specifically:
The tariff-cartel dumping systems of America and Germany operated in a way that lowered American and German unit costs and raised British unit costs, facilitating constant undercutting of British prices and erosion of British marketshare.
The short-run cost disadvantage of the British mills was progressively translated into a long-run loss of competitiveness as American and German mills maintained higher levels of capital investment. Dumping affected this process directly by increasing British investment risk and diminishing American and German investment risk.
Dumping Lowers Cost
The aspect of dumping that most engaged the attention of contemporary steel producers was its effects on the relative unit costs of dumping firms, on the one hand, and of firms in whose markets dumping was occurring, on the other hand. It was the consensus of British, American, and German industrialists that dumping lowered the unit costs of the dumpers and raised the unit costs of the "dumpees." The reason was quite simple—the dumpers could engage in the practice known then as "continuous running" or "rapid driving," that is, running their mills at high operating rates, which resulted in progressively lower per unit pro-
duction costs for each additional unit of output. Because of high tariffs, they could dispose of surpluses abroad without spoiling domestic prices, and in fact could maintain high domestic prices by limiting the available supply within the home market. British "dumpees" generally could not do this; foreign markets were increasingly closed to their exports, and continuous running for purposes of serving only the home market tended simply to further depress prices in that market, without necessarily increasing sales volume.34 "Continuous running" had a particularly dramatic impact in capital-intensive industries with high fixed costs, that is, sunk costs incurred whether or not goods were actually produced. In such industries (steel, chemicals, machinery) in competition between two facilities of equal efficiency, the producer able to operate at the highest rate would enjoy the lower costs, and hence, the competitive edge. The most modern and efficient production equipment carried with it very high fixed costs. While such facilities, if run "flat out," could produce goods at a lower cost per unit than those of any competitor, if the utilization rate dropped and the works were put on halftime or quarter-time, the cost per unit could easily be higher than that of older, less efficient facilities.35 This dynamic could thus enable a less efficient producer to operate at lower unit costs than a more efficient firm.
The ability to run plants "flat out" was one of the principal policy justifications offered for a high tariff and the "trust" form of combination that was offered by Carnegie, Schwab, and other U.S. industrialists of the time.36 The German
Stählwerksverband (steel cartel) likewise justified its dumping policy primarily in terms of the beneficial effect on operating rates.37 For the British, the inability to match the American-German practice of ''continuous running" produced a vicious circle—loss of sales volume effectively raised British mills' unit costs, making them vulnerable to still further loss of sales and additional increases in costs.38 A British scholar, noting this phenomenon, observes that
The importance of 'dumping' in any explanation of Britain's difficulties [in steel] may well have been overstated in the Report of the Tariff Commission, but there is no question that selling below average cost gave both German and American exporters a very real competitive advantage in world markets. The constant refrain of witnesses before the 1904 Tariff Commission was that British manufacturers were inhibited from pursuing a like policy because of 'the openness' of the home market, in which American, German, and Belgian manufacturers were making growing inroads by 'unfairly' undercutting domestic producers. If the home market—relatively stagnant though it was—wasn't reserved for native manufacturers, it was no wonder that they were slowly demoralized by being placed in a disadvantageous position in overseas trade.39
Dumping and Capital Investment
In 1870 Britain possessed the most modern, competitive iron and steel industry in the world, but by 1914 it had fallen behind its rivals by virtually every standard used to measure international competitiveness. 40 Subsequent generations of scholars have concluded that a decisive element in this competitive rever-
sal was the fact that the British iron and steel industry did not make the capital investments prior to 1914 that would have enabled it to remain competitive in subsequent decades.41
In fact, at the turn of the century British steelmakers were quite aware that their competitive edge was slipping away due to their own failure to invest in state-of-the-art technology, and they said that dumping was a principal cause of that failure. One of them testified as follows in 1904:
I state emphatically that, in my opinion, some fiscal provision for meeting dumping is essential to the maintenance of the trade of this country. . . . I may explain the above statement further—continual changes and improvements are being made in the manufacture of steel. I know of no trade which of late years has been subject to so many changes and improvements in the mode of manufacture. All these changes involve enormous outlays. If the manufacturers in this country are unable, from the instability of their market and from the liability of being overwhelmed from the stuff being dumped upon them at prices with which neither they nor anybody else can compete, if they are deterred from making the necessary improvements from availing themselves of new inventions, and if the foreigners, by reason of their protected market and of the certain large returns which they get from their home trade, are enabled to make these outlays, they will place themselves in such a position that, even if we got a free market and a free interchange, we should be some years before we could overcome them. I fear that if this state of things goes on for a considerable time longer, we never shall get on equal terms; they will attain so much superiority, and, when we are driven out of the market, our competitors would raise their prices to us.42
For British entrepreneurs, the decision as to whether to invest in the latest steelmaking technologies turned, in large degree, on their assessment—and that of the capital markets—of the risks involved. Although large new mills could produce goods at lower average costs than a collection of smaller facilities, because of the higher fixed costs of the biggest facility, its losses were potentially larger if demand fell and it could not be utilized adequately. If demand were sharply and continuously cyclical, the biggest facility might not have the lowest average costs even over the long run because its unit fixed costs would be so high in each recession. Alternatively, if cyclicality were somewhat less sharp, the biggest facility might suffer higher losses in recessions, but enjoy lower average costs over the long run, and thus be more profitable.43 Whether this prospect was sufficient to entice a given entrepreneur into committing the massive sums of capital needed to establish the largest facilities depended on the entrepreneur's assessment of the risks involved:
[T]he rational choice depends upon the entrepreneur's attitude toward risk and his ability to insure himself against it. In either case greater variability in demand can make it rational to invest in less capital-intensive plants, even if those plants are less efficient in terms of minimum average cost. Only if demand fluctuations were reduced would the most efficient, capital-intensive technologies always be the most attractive.44
Dumping—or rather the protection/cartel system to which dumping was integral—affected this equation in several ways. First, for the Germans and Americans, by reducing competition and enhancing profitability in the home market, dumping diminished cyclicality and reduced investment risk, making it less hazardous for entrepreneurs to invest in the most advanced capital facilities.45 Many German scholars concluded that the tariff-cartel system enabled German industry to achieve higher levels of technological advance and production efficiency than could have occurred under laissez faire:46
The tariff-cartel system could keep domestic prices high enough to cover fixed costs, while firms added to their profits by selling at marginal cost on the depressed world market. . . . In ten of the years between 1876 and 1896 the average cost (Selbskosten) of rails from the Krupp firm exceeded the average export price, the world price with which the British had to contend. The domestic price, on the other hand, always exceeded average cost by over 10 marks per ton. Thus, the tariff-cartel system kept the German mills like Krupp profitable through times of recession. The national statistics also reflect such a pattern. In 1886, 1891, 1901, and 1908, when international trade crisis struck, British pig iron output declined 10 percent on average from the previous year, and the number of furnaces in blast fell 14 percent. German pig iron output fell only 6 percent, and the number of furnaces in blast dropped only 5 percent. Given the greater riskiness of their market environment, British steelmakers may have been rational to use less capital-intensive techniques, even if that meant slightly higher average costs.47
Second, the British confronted not only the loss of sales in protected foreign markets, but more violent cyclical swings in their own market as a result of intermittent incursions of dumped products, which exacerbated the intensity of recessions. This was the nub of the British strategic dilemma—investment risk was higher for them than for the Germans or the Americans. "In England there is not the same security for capital," as one steel tubemaker put it.48 Prime Minister Balfour explained the problem at the height of the British trade policy debate in 1903:
Now, there is no reason to expect that the plant erected to meet an average demand would reach the exact size most conducive to economy of manufacture. . . . Neither is it practicable to arrange that the plant shall always be kept working full time. If it is, there must evidently be recurrent period, during which over-production . . . must inevitably take place. Such is the ordinary position of the manufacturer under free trade. Compare it with the position of his protected rival, who controls his home markets. He is not haunted by the fear of overproduction. . . . [S]o long as other countries are good enough to offer him open markets, he can dispose of his surplus abroad, at prices no doubt lower, often
very much lower, than the price which his quasi-monopoly enables him to obtain at home, but at prices which nevertheless make the double transaction, domestic and foreign, remunerative as a whole. . . . The manufacturing capitalist [in the free trade country], when investing his money in costly plants has, in any case, many risks to run—new inventions, new discoveries, new fashions. Add to these his loss, actual or anticipated, through the operation of foreign protection, and his burden becomes insensibly increased. But add yet again the further uncertainty and the further loss due to the system [of protection/cartels/dumping] and he is overweighed indeed. Will the hostile combination keep together long enough to ruin him? Can his credit stand the strain? Is it worthwhile holding on in the face of certain loss and possible ruin? These are questions which the lenders of the threatened industry cannot but ask. And surely the mere fact that they have to be asked must shatter the buoyant energy which is the very soul of successful enterprise.49
Balfour's diagnosis was corroborated by the contemporary testimony of many British steelmakers.50 But the cure was never found. No action was taken to curtail dumping or to open the foreign markets from which dumping was occurring. The immediate cost disadvantages and loss of sales confronting British steelmakers as a result of dumping were gradually translated into a loss of competitiveness that would, in the years to come, prove to be irremediable.
Dumping and Consuming Industries
During the British dumping debate, free trade advocates argued persuasively that dumping of intermediate products had actually enhanced the international competitiveness of British industries that used those products as inputs. Thus, although dumped sugar may have weakened England's sugar refining industry, cheap imported sugar fostered new food processing industries—jam, confectionery, biscuits, condensed milk—that employed far more people than the sugar refining industry had ever utilized. British shipbuilders reported that they bought dumped German castings and forgings, "built them into ships and machines, and sent them back to Germany."51 Even within the British iron and steel industry
itself—the principal source of alarm over dumping—many manufacturers benefited, at least in the short run, by purchasing low-priced German and American iron and semifinished steel and said as much.52 Dumped imported steel not only fostered a price advantage, but occasionally offset domestic shortages or attempts by domestic iron and semifinished producers to "boycott" downstream steelmakers.53 But even the firms that gained apparent short-run advantages from buying dumped inputs expressed misgivings about their growing dependency for inputs on foreign syndicates that were beginning to move into areas further downstream in the production process. Switching to a foreign source was not always a discretionary option; the price advantages associated with buying dumped inputs in some cases forced British manufacturers to abandon their own internal production of those inputs54 and in other cases, to switch from domestic suppliers to foreign sources so as to remain competitive, despite certain troubling long-run implications. 55 Reflecting this concern, a pattern that emerges from contempo-
rary surveys is the intermittent and unpredictable character of the German and American sales; they occurred in waves that surged and receded, discomfiting competing British producers in the former case and British consuming industries in the latter.56
German steelmakers recognized that their dumping abroad created serious competitive problems for their domestic steel-consuming customers, who had to compete with foreign firms that were able to buy dumped German steel. Thus, for example, Dutch producers of barges for river transport were able to capture most of this business from their German rivals at the end of the nineteenth century because they could buy dumped German plate, while their German competitors paid the higher domestic price set by the German steel cartel.57 The German syndicates counteracted this problem, however, by developing a sophisticated system of "export bonuses" or "bonifications"—payments to domestic customers who could demonstrate that their inputs would be utilized for producing products for export.58 This system underwent continual refinement that served to strengthen the vertical and horizontal cohesion of German industry even as German dumping was weakening and in some cases breaking the vertical relationships between British producers. Thus, German export bounties were paid only to downstream exporters who brought exclusively from the upstream cartels, and only downstream firms that were themselves members of cartels in their own industries could qualify for export premiums from upstream cartels.
With the advent of the system of export bonuses, the focus of German dumping began shifting from intermediate industrial products toward finished products and higher value-added products. The Germans found that under the new system, the purchasing power of the home market was increased because downstream firms could, in effect, buy semifinished materials for use in production for export at "dumping" prices and expand their export sales of finished products through
price reductions. The U.S. Consul General in Berlin, commenting on this phenomenon in 1916, observed that through the new selling policy
[t]he Steel Verband therefore shifted the entire movement of half-finished materials, and the ultimate result was the struggle for supremacy in finished manufactured products in South America and the colonies.59
A number of British manufacturers abandoned intermediate production processes in the face of German dumping, retreating into production activities further "downstream," in many cases becoming dependent on dumped German products as inputs. This was rationalized on the grounds that the downstream product areas were more specialized and remunerative, and that it therefore made sense to allow the Germans to supply the commodity-grade inputs while concentrating on areas requiring the greatest craftsmanship. But British industrialists soon found that they had not only lost control of the upstream part of the production process, but confronted German dumping in the downstream product lines as well. One British steelmaker testified in 1904:
Our profits dropped from £30,000 in 1899, to about £2,600 in 1900. The following year, there was a loss of nearly £10,000, and it became obvious to those interested, that the cause of this sudden and alarming change in the prosperity was due to the heavy dumping of steel from Germany and America, at a price sold in this country considerably below the cost price at which the steel could be made here. . . . The alternatives we had before us were either to wait in the hope of a change in the conditions of trade or to put down fresh plant and get into a higher class of manufacture. We settled to do the latter, and by large expenditure gradually got into a different trade. . . . Then the proceeding was this. First, we began to make rails, sheet bars, & c., then the Germans, by dumping, stopped the trade in bars, blooms and billets. We then went into the tram rails and electric rails, and they are now beginning to dump those articles; finally, we were driven into a general trade in which we also suffered from German competition. There is no part of our trade in which they are not to some extent dumping, only this is not so acute as when we were confined to the lower grade of material worth from £3 15s to £5 per ton. . . . Then again, if we get into a still higher class of trade where wages per ton are greater still, they cannot touch us—at all events at present—but I think they are bound in time to touch us in higher products. We are beginning to feel it already.60
National Security Implications
British imperialists noted that dumping was destroying certain "primary" and "staple" industries and warned that foreign competition could weaken or eliminate industries that were essential to national security. The Free Traders ques
tioned whether, in light of their poor competitive performance, such industries were, in any event, "one[s] which we could expect to keep." The loss of "staple" industries was more than offset, they argued, by the advent of specialized downstream industries processing the cheaper imported inputs formerly produced by the supposedly vital basic industries.61 Others made the point that rapid strides in human technological progress and global economic and financial integration had made war virtually unthinkable; the advent of technological breakthroughs in communications and transportation and growing foreign direct investment were bringing about the very international "division of labor" envisioned by Adam Smith. Commenting on globalization in 1910, one observer noted that there now existed
a financial interdependence of the capitals of the world so complex that disturbance in New York involves financial and commercial disturbance in London, and, if sufficiently grave, compels financiers of London to co-operate with those of New York to put an end to the crisis, not as a matter of altruism, but as a matter of commercial self-protection. . . . [T]his complexity of the international division of labor tends to render futile the . . . contrivances of conquest. . . .62
The Surprise of 1914
Such ruminations were abruptly removed from the realm of abstract argument on the morning of August 4, 1914, when thousands of German troops crossed the Belgian frontier and began an assault on the fortress city of Liège, the beginning of an onslaught by 1.5 million men against the eastern boundaries of Belgium and France. A few hours later, to the considerable surprise, if not outright disbelief, of most of its citizens, Britain found itself at war with Imperial Germany. Five days later, on August 9, 1914, 80,000 men of the British Expeditionary Force, representing virtually all of Britain's professional army that could be gathered in the home islands, began embarking for the continent, and within days its "tiny numbers were sucked inexorably into the military planning of the great continental powers."63 On August 13, 1914, the Germans began bombarding the forts defending Liège—which had been expected to hold out for months—with terrifying new weapons that the British did not know existed—huge 420-mm howitzers, the "Big Berthas"—that had been developed secretly at the Krupp steelworks in Essen. Smashed to pieces by these guns, all of the forts fell within four days; the German armies passed through Liège and began a sweep across Belgium. Brussels fell on August 20. On August 21, 400 German guns, include
ing the Big Berthas, began bombarding the Belgian fortress city of Namur; the city fell two days later, unhinging the whole Franco-Belgian line and forcing the French army into a general retreat along its entire front.64 The same day, a little further to the west, the British Expeditionary Force, which had advanced into Belgium and taken up positions along the Mons Canal, came under attack by greatly superior German forces. Outnumbered, out of contact with the French, and threatened with envelopment on both flanks, the British began a fighting retreat that did not end until they reached the Aisne River, 250 miles to the south.
The stunning events of this fortnight in August 1914, which catapulted Britain into a major war on the continent and saw the unravelling of pre-war allied strategy, brought only the first of a succession of unpleasant surprises to the nation. The military had expected that if war came, it would be short and sharp, resolved with a few decisive battles, such as the Franco-Prussian contest of 1870. British marksmanship, pluck, and military professionalism would carry the day. There was no reserve army to call up, no store of munitions to sustain a long war, and no arrangement for industrial production to support a continental-scale army in the field.65 Instead, after the first engagements in Belgium and Northern France, the war degenerated into the ghastly deadlock of trench warfare, in which, as was quickly demonstrated, sheer weight of munitions and numbers counted most. 66 Less than three months after the outbreak of the war, the Germans launched a massive attack on British positions in front of Ypres, concentrating what was at that time the greatest weight of artillery that had ever been brought to bear on a British force in the field and mounting successive, massive infantry assaults with a numerical superiority of between four and seven to one. This continued for four weeks. The British held their ground, but by mid-November, 1914, the original British Expeditionary Force had largely ceased to exist; a third
of its original members had been killed outright, and many others were wounded; ''the British regular army no longer had the capacity to fight a major battle."67 Thereafter, Britain, like the other belligerents, had no choice but to mobilize its civilian population and its economy to fight a protracted war of attrition.
The unexpected development of a stalemate on the Western Front brought in its wake other, even more fundamental surprises. One was the extent to which Britain's principal strategic asset, its fleet, was neutralized by German industrial power. The British blockade against Napoleon had played an important role in bringing him down; but the British blockade against Imperial Germany was countered, to a considerable degree, by German industrial science and technology. German manufacturing concerns quickly converted to the mass production of munitions, spewing out guns, shells, and bullets at an incredible rate.68 Chemical companies churned out not only high explosives but a vast range of ersatz products to replace items that had been cut off by the British blockade. German railroads shifted huge armies rapidly around the interior of Europe, whereas the British fleet operated more or less ineffectually around the periphery. Indeed, Britain soon found itself under partial blockade as German U-boats began sinking the merchant ships that constituted the country's lifeline.
But the most appalling surprise—known to Britain's leaders but not the public—was the sheer extent to which the country's industrial base had decayed. In the wars against Napoleon, the "workshop of the world" had outfitted not only the British fleet and army, but also the large armies of its continental allies—Prussia, Russia, Austria, and Spain. But, by 1914, so many industries had disappeared or fallen behind technologically that Britain could not sustain her own army and navy, much less those of her allies:
[T]he first two years of the Great War showed England to be incapable of fighting a major war from her own industrial resources.69
The Ammunition Shortage
An early and continuing manifestation of British industrial weakness was the ammunition shortage, a scandal that erupted several months after the outbreak of the war. In the midst of the Ypres battle, the British commander on the scene warned London that unless he received large quantities of ammunition, he would
be forced to fight without artillery support. He was told to economize. 70 Several months later, in March 1915, Douglas Haig, in command of Britain's First Army in Flanders, proposed a major assault on German positions along Aubers Ridge, but there was a problem:
I went out to Hazebrouck about 10:00 am and saw Sir John French. He approved my plan of operations but there was no ammunition. . . . This lack of ammunition seems serious. It effectually prevents us from profiting by our recent success and pressing the enemy before he can reorganize and strengthen his position [original emphasis] 71
The British army lacked, in particular, large numbers of high-explosive artillery shells of the type needed to make an impression on the Germans' well-engineered and deeply dug trench systems; for the most part, the British possessed only airburst shrapnel shells that made little impression on an entrenched foe. Over the next four years, the British army paid dearly for this deficiency, which was never wholly made good; the shell shortage limited the army's ability to sustain offensive action, or, if an attack was made, greatly increased the cost to the attackers, since the German trench systems were seldom adequately softened up by preliminary artillery fire.72 On May 15, 1915, a British offensive at Festubert was broken off because the British forces had expended their ammunition along the entire front. 73 In the fall of 1915, compelled to take the attack at Loos to relieve German pressure on their Russian allies, the British assaulted the German trenches despite the fact that the shell shortage left them without adequate artillery support—"all we wanted was ammunition." They suffered 60,000 casualties, making no appreciable dent in the German lines.74 In Britain the shell shortage fueled
public outrage, and much angry finger-pointing took place over who bore the blame; partly as a result, Prime Minister Asquith was compelled to ask the resignations of all of his ministers in 1915 and to form a coalition government.75
The Steel Shortage
The ammunition shortage was, in part, a function of poor planning and bureaucratic bungling, but it was much more fundamentally a reflection of the fact that Britain's industrial base could not meet the demands that were being placed on it. It was evident in 1915 that the British steel industry could not even begin to produce enough shell-quality steel; the three firms that could make such steel had a combined output of 5,000 tons per week, while the government was asking for 35,700 tons per week.76 The shortfall was slightly alleviated by reducing the quality requirements for shell steel and by attempts to import shell steel from the United States, but Britain's shortage of shell steel "remained acute until the end of the war."77 In 1918 Britain's ability to go on the offensive on the Western Front was still constrained by a shortage of artillery shells, with respect to which "steel [was] the limiting factor."78 The Ministry of Munitions concluded at war' s end that
It was only the ability of the Allies to import shell and shell steel from America and iron ore from neutral Spain that averted the decisive victory of the enemy.79
The ammunition shortage was only one symptom of a broader problem, the inability of the nation's steel industry to produce the quantity and quality of steel needed by the nation's armed forces to fight the war. It was a "steel war," in the words of Sir Winston Churchill, the Minister of Munitions. Steel was needed, most critically, to produce the merchant ships that constituted Britain's supply lifeline; after that, steel was needed for naval vessels, shells, artillery pieces, rails, construction of fortifications, and later, for tanks. The demand could not be met
from the domestic production base that was, moreover, heavily dependent on imported ore and semifinished steel that had to be brought in through U-boat-infested waters. The Germans identified Britain's steel dependency as one of its greatest strategic vulnerabilities and made the overseas steel lifeline the primary target of its policy of unrestricted submarine warfare.80 Moreover, even leaving aside the U-boat problem, the overseas sources of ore and semifinished and finished steel were precarious at best. Germany, the principal source of imported semifinished steel, was now the enemy; Belgium, another source of steel, was occupied, as were most of the iron ore fields of France.8l Britain's allies, Italy and France, were utterly unable to meet their own steel needs and looked to Britain to do so. Sweden, a primary source of iron ore, was an unfriendly neutral. The United States proved to be an important, but erratic source of supply.82
These problems were never overcome. The shortage of steel meant that it had to be rationed between competing demands; Churchill identified steel as one of the four limiting factors of production (the others being labor, shipping, and money) confronting the Ministry of Munitions. 83 The country's annual wartime needs were estimated at 10 million tons; it could only produce 8.5 million tons, and "out of this, every requirement must be met, and if through shipping shrinkage the total production is reduced, all programmes will be affected."84 Because of the shortage, a constant tug of war raged within the British war effort to secure supplies of steel. In 1917, Churchill wrote to Haig and observed that
there are many difficulties here, both with labour and materials, especially steel, and at this stage of the war, it will often become necessary to choose between desirable things and to throw special emphasis on this or that branch of production.85
Britain, which a few years earlier had debated whether it should worry at all about the erosion of its steel industry, now frantically attempted to expand that industry and make good the many deficiencies that the war had revealed. "It was . . . urgently necessary to increase the capacity of the British steel works at once." The government appealed to the steelmakers to expand their capacity, but the British mills replied that capital expenditures on the scale needed were too risky. They expressed concern about the international competitive environment they would face after the war. The government advanced much of the capital required, but the expansions of existing mills proceeded much more slowly than hoped, due to the recalcitrance of the steelmakers to expand and shortages of supplies and labor, which were "urgently needed for war purposes." Most of the steel works "extensions" were still incomplete when the war ended in 1918.86
The Sluggish Production of Tanks
Because the steel shortage touched every aspect of Britain's war effort, major and minor, it is impossible to assess its full effect. In a myriad of specific cases, the shortage meant that there was not enough merchant tonnage, not enough shells, or not enough artillery pieces to perform the task at hand.87 Often, the deficit was made up in additional lives lost, the most dramatic example of which was Britain's sluggish deployment of tanks, the weapon that ultimately broke the deadlock in the trenches. The idea of armored caterpillar-tracked vehicles as a way out of the trench stalemate was conceived at the very beginning of the trench deadlock in late 1914 and early 1915. 88 But it took over three years to develop this idea from a prototype stage to that of a weapon capable of being employed on a mass basis in the field; only small batches of tanks were available before late 1917, not enough to achieve any decisive result. During this interval, from early 1915 to late 1917, the British army launched repeated mass infantry offensives against the German trenches at the Somme (1916) and Passchendaele (1917), suffering truly horrific losses without achieving any significant result.89
Britain's tardiness in employing tanks on a mass scale was due to several
factors, including resistance by some elements of the military, labor shortages, training bottlenecks, and disputes over design issues. 90 But material shortages and shortcomings played a crucial limiting role. The Admiralty engaged in a continuous struggle with the tank program over the supply of steel plates, and the Admiralty usually came out on top.91 Perhaps more seriously, the British steel industry experienced great difficulty producing adequate numbers of castings that would join together the links of the tanks' caterpillar tracks. Until mid-1917 British steel manufacturers could not make adequate numbers of track links at the tensile strength required. As a result,
the production of track links proved a limiting factor in the output of tanks and contributed considerably to the accumulation of arrears, since it was always some six months from the time of placing a contract before a new foundry could produce satisfactory links in considerable quantities.92
What might have been achieved, and the losses that could have been avoided, had tanks been employed en masse at an earlier date, was revealed when the first mass tank assault was undertaken by the British at Cambrai on November 20, 1917, in the last year of the war. Four hundred seventy-six tanks, backed by infantry, broke a six-mile wide hole through the vaunted German Hindenberg Line, penetrating four and one-half miles through the German positions in a single day. Winston Churchill commented as follows:
Accusing as I do without exception all the great ally offensives of 1915, 1916 and 1917, as needless and wrongly conceived operations of infinite cost, I am bound to reply to the question, What else could be done? And I answer it, pointing to the Battle of Cambrai, 'This could have been done.' This in many variants, this in larger and better forms ought to have been done. . . .93
A later author points out that Churchill was wrong insofar as he implied British leaders were at fault for the delay; the real problem was with the production base:
[O]f greatest significance was the fact that only now [late 1917] had tank production reached a level at which this weapon had the weight of numbers largely to affect a military operation. This matter of numbers, depending on the painful development of a productive capacity, needs to be stressed.94
See, generally, Swinton (1972), op. cit.; History of the Ministry of Munitions (1922), op. cit., Vol. XII, Part III.
"Each of the armed services was fighting to obtain every ton of steel and freight for itself in a freefor-all in which the Royal Navy seemed to come off best" (Bryan Cooper, The Battle of Cambrai [New York: Stein and Day, 1967], pp. 52-53).
The problem was "blowholes and segregation of sulphur and phosphorus at the parts of the link where soundness was particularly essential" (History of the Ministry of Munitions , op. cit., Vol. XII, p. 48).
Churchill (1923), op. cit., Vol. IV, p. 61.
Wilson (1986), op. cit., p. 487.
Consequences of a Weakened Industrial Base
Although the problems of the steel industry posed the greatest difficulties for Britain in World War I, the nation's industrial shortcomings went far beyond this industry or any of the industries normally associated with munitions production. The domestic industrial base producing consumer goods had also eroded, and therefore could not be converted to war production on the scale required. Before the war, Britain's clockmaking and mechanical toy industries had been displaced by imported clocks and toys; when war came there was no precision clock or toymaking industry that could be converted to the production of accurate shell fuses. Britain had to create a light engineering industry to fill this gap, but the effort to do so revealed another hole in the nation's industrial fabric: the lack of a modern machine-tool industry that could make the machines needed to run the production lines. The nation had become dependent on imported machines, and only the importation of machine tools from America, Switzerland, and Sweden "prevented a total breakdown of the British effort to create new industries between 1914 and 1916."95
Britain had become dependent on Germany, now her enemy, for many of the industrial products needed to wage a modern war. At the outbreak of the war, Britain looked to Germany for 90 percent of the optical glass used for precision instruments, for 75 percent of its glass for electric lights, and even most of the laboratory instruments used by British scientists. Britain was dependent on Germany for chemicals needed for explosives and even drugs such as aspirin; during the war it "had no alternative but to continue importing German drugs via neutral countries." Britain was dependent on Germany for precision bearings and magnetos, both of which were indispensable for every type of motorized equipment—aircraft, tanks, trucks, cars. Britain tried, but never succeeded, in making up for this shortcoming by imports from Sweden and Switzerland; in fact, Britain was never able to produce the number of engines needed to fight the war and had to rely on its allies to make good at least part of the difference.96
Although the decline of Britain's strategic industrial base that had occurred by 1914 had multiple causes, dumping was an important contributing factor. The stagnation of investment by the British steelmakers in the years prior to the war, and indeed, through the war years themselves, reflected the demoralization that had set in as the British confronted a competitive dilemma for which they had no solution. While it is impossible to state with certainty that the qualitative problems revealed by the war, such as the difficulties in making track links for tanks, could have been resolved more readily by a larger and more robust steel industry, it is reasonable to assume that a bigger and a more vibrant industry would have grappled with such challenges more successfully.
Britain's experience allowing unrestricted dumping a century ago is obviously subject to varying interpretations, but on balance its experience can hardly be held up as a ringing testimonial to the wholly passive policy that was actually followed. Dumping in the markets of the British Empire by cartels operating behind high tariff walls placed a number of key British industrial sectors at a permanent competitive disadvantage. Over time, this led to declining relative competitiveness and, ultimately, disinvestment. This erosion of the British industry left the Empire dangerously vulnerable when it was unexpectedly plunged into a major war in 1914. Although it was argued contemporaneously that dumping of industrial inputs such as steel enhanced the competitiveness of downstream industries that consumed these inputs, it was also noted at the time that the downstream industries themselves suffered from the erratic availability of dumped inputs, the growing dependency on their direct competitors for key inputs, and ultimately by dumping in their own downstream markets—and it was not merely the steel industry, but British industry as a whole that had declined dramatically by 1914. Finally, unrestricted dumping surely gave rise to short-run benefits to consumers of dumped products, a fact that was recognized and played a major role in the electoral victory of the Free Traders in 1906. However, seven hundred thousand of those same British "consumers" were killed several years later in the war, and millions more wounded, a toll far higher than it would have been but for the erosion of Britain's industrial base.
DUMPING TODAY: STILL A PROBLEM?
In the decades following World War I, many countries enacted antidumping rules that provided for the imposition of duties at the border on imports that were being sold at "less than fair value" and injured a domestic industry.97 The "problem" of dumping was recognized in many bilateral trade agreements in which treaty partners agreed to the mutual use of antidumping measures to offset dumping. When the GATT was negotiated in 1947, Article VI provided that contracting parties could use antidumping duties to offset dumping that caused material injury. In subsequent rounds of multilateral trade negotiations, the contracting parties have adopted and refined a succession of antidumping codes that prescribe detailed procedural and methodological rules for the application of antidumping duties.
Notwithstanding the proliferation and refinement of antidumping rules, antidumping measures remain controversial. As a leading GATT scholar observes, "central to the whole subject is the perplexing question whether antidumping law and policy, as related to international trade, makes any real policy sense today at all."98 A chorus of academic commentary answers that question in the negative.
The British experience at the turn of the century would suggest that unrestricted dumping, at least in that era, was harmful. The question remains whether the events of that earlier era have any relevance today. It might well be argued that the widespread adoption of liberal trade and competition policies worldwide has changed the international commercial environment so profoundly that the dilemma which Britain once confronted does not and could not exist today. In addition, it may be that technological change has rendered obsolete the competitive dynamics that existed at the turn of the century. In fact however, while the world has undergone revolutionary political and technological change since that time, the factors that made dumping harmful remain with us today.
Dumping and Cartels: the Evolution of Trade and Competition Regimes
Superficially little remains of the international trading order that existed in the early twentieth century. The world is no longer clearly bifurcated between a single, major free-trading empire and a group of protectionist states and empires. Since the inception of the GATT in the late 1940s, quantitative import restraints have been phased out and tariff walls have been progressively dismantled. In the United States, an antitrust movement fostered enactment of strong antitrust laws and the breakup of many American business trusts. Following World War II, U.S. antitrust thinking was widely (albeit not universally) embraced abroad and some large industrial groups were dismantled by U.S. occupation authorities in Germany and Japan. National competition authorities were established, under U.S. prodding and tutelage, to curtail anticompetitive business behavior. On the basis of such changes, many observers concluded that the types of problems once posed by German cartels and American trusts had become marginal issues in the world economy.
In fact, a significant portion of the debate over the need for antidumping measures turns on the question whether the market barriers and cartels of an earlier time have really been banished from the world economy to the extent assumed by many, or whether they have simply been driven underground by the evolution of national and multilateral competition rules. Numerous critics of antidumping measures concede that dumping from protected "sanctuary" mar-
kets by cartels or monopolies may well be a potential problem, were it to occur, but that such practices are rare today:
The world has changed somewhat in the last eighty years. The operations of foreign cartels are likely to be a smaller problem today than in 1900, in part, at least, because of the operation of national competition laws. Accordingly, even if a persuasive argument that dumping of this type creates economic losses for the receiver of the dumped goods was available, the relevance of the argument to the modern world might be doubted.99
There is no question that markets are more open today and that restrictive private arrangements have much less effect on trade today than was the case in the first half of this century. However, the dismantling of tariff walls has not necessarily resulted in open markets, and it does not follow that because national antitrust laws and enforcement agencies now exist, anticompetitive business behavior has ceased or been reduced to marginal importance.l00
Although "national competition laws" are cited reassuringly by antidumping's critics as a latter-day bulwark against cartels, the assumption that such laws have largely eliminated cartels does not withstand scrutiny. It is true that following World War II many countries enacted competition laws that incorporated elements of U.S. antitrust doctrine, including proscriptions against price fixing, joint restraints on output, and the like. But these laws were grafted onto political systems whose industrial traditions were quite different from those of the United States, and in which antipathy toward joint industrial action, and even cartels, was far less. Following the immediate postwar era—a sort of high water mark for U.S. antitrust ideals—some national competition laws were amended to permit the formation of cartels as industrial policy tools. In other cases the laws were simply not enforced, with governments either tacitly consenting to cartel activity or, in some cases, actually encouraging and directing it.10l The U.S. government gradually acceded to such arrangements, initially reflecting practical foreign policy considerations (e.g., the onset of the Cold War) but increasingly because it lacked the ability to do anything about the problem without incurring
political costs that were unacceptably high. 102Over time, a rough modus vivendi emerged in which many countries' governments and competition authorities paid obeisance to U.S.-type antitrust principles, while at the same time cartels and cartel-type arrangements were allowed to regulate a substantial volume of world trade.103
THE FACTUAL DEFICIT
Public discussion of the persistence of a ''cartel problem" is hampered by the deficit of readily available hard information on the subject. Although at one time the U.S. government and the academic community possessed a wealth of information about national and international cartels, that information base has largely disappeared. The U.S. government agencies that once gathered information on this subject no longer do so and do not possess the capability that once existed. The U.S. economics profession has largely abandoned empirical study of such practices in favor of theoretical mathematical modeling exercises, and this approach has come to dominate the thinking of U.S. antitrust enforcement agencies. As a result, apart from the piecemeal facts that may emerge from a particular episode of civil litigation or as a result of sectoral studies by individual scholars, very little of a practical nature is known about the extent and pervasiveness of restrictive business practices outside of the borders of the United States.
The factual void with respect to anticompetitive business practices in international markets plays a central role in the present debate over antidumping. Critics of antidumping policy are able to ignore or downplay the notion that cartels, monopolies, and similar entities exist abroad that foster dumping because there is little empirical information on the subject that offers a contrary perspective. The result, typically, is a steady stream of highly visible, authoritative pronouncements by prominent scholars on the subject of dumping, which, when juxtaposed against market realities, serves simply to underscore the surreal character of the debate.
Several examples of how the factual deficit colors the antidumping debate were provided by the investigation conducted by the U.S. International Trade Commission in 1994-1995 with respect to the rationale for antidumping and countervailing duty laws.104 The Commission's staff chose to approach this question through a mathematical modeling exercise, and its factual inquiry was largely limited to the gathering of quantitative duty for use in its models. The existence, extent, and implications of anticompetitive combinations abroad giving rise to dumping was not considered relevant to the exercise. A number of prominent academic witnesses who testified similarly ignored or downplayed the role of anticompetitive practices in fostering dumping. Based on this approach, the academic witnesses and the Commission staff tended to reach similar conclusions—that antidumping measures are unwarranted, at least in most cases. But a fleshing out of the facts surrounding cases presented in the investigation calls into question the soundness of the conclusions reached.105
One prominent witness before the U.S. International Trade Commission, Professor Robert Willig of Princeton, concluded that so-called "strategic dumping," involving aggressive pricing of exports in combination with protection of the home markets of the exporter, "may, in the long run, harm consumers in the country that receives the exports," as well as domestic industries in those countries. So-called predatory dumping (designed to drive competitors out of business) posed similar concerns. That said, however, he reported that a study conducted by one of his colleagues, Dr. Hyun Ju Shin, had concluded that almost all recent U.S. antidumping cases with a non-negative outcome did not involve either strategic dumping or predatory dumping. The approach suggested by Willig was sensible—that is, to examine the central question of how truly prevalent are the anticompetitive arrangements that supposedly give rise to problematic dumping. However, the examination itself was flawed because of the lack of factual information available. The conclusion was based on Dr. Shin's assumptions rather than on empirical study.
Dr. Shin examined 282 non-negative outcome antidumping cases, "screened out" all cases that did not involve a threat to competition in the United States (e.g., "strategic" or "predatory'' dumping), and found that only 20 to 30 cases
involved actual threats to competition in the United States. This would appear to be a fairly serious indictment of the entire antidumping regime. But the standards that Dr. Shin used to eliminate cases were not based on empirical sectoral case studies, but on her own assumptions concerning what seemed logical to include or exclude (e.g., "fairly sensible screens"). Although her reasoning is quite defensible, in an abstract sense, at least one of her most important assumptions was sharply at odds with commercial reality. She surmised that
[i]f the imports that were challenged were coming from five or more different countries, then it seemed implausible that even if U.S. suppliers were eliminated from the market place, that importers from five different countries could form a cartel—although I am interested in checking this out with other panelists—and thereby participate in a monopolizing episode of the U.S. market.106
The assumption that producers from five or more countries could not form a cartel is inaccurate with respect to most cases involving steel products, which in turn comprise the largest single group of antidumping cases.l07 It is also erroneous with respect to a number of other industries.108 It is, in other words, a factual error of sufficient magnitude to throw into question the validity of the study's
basic conclusion, which is that truly problematic dumping (strategic) is relatively rare.
Another example is offered by the work of one of antidumping's harshest critics, J. Michael Finger of The World Bank, who also testified in the 1994 U.S. International Trade Commission proceeding. His book, Antidumping: How It Works and Who Gets Hurt, denounces antidumping measures based on a series of case studies of the application of U.S. antidumping measures.109 The case studies are less interesting for what they contain—a series of accounts of the difficulties experienced by foreign firms subject to antidumping measures—than for what is omitted, that is, any description whatsoever of the cartelized milieu out of which dumping arises.
One of the centerpieces of Finger' s book is the Swedish stainless steel industry. The author of this case study, which is entitled "Antidumping Attacks Responsible International Citizenship," portrays the Swedish specialty steel industry in glowing terms as "an industry following good economic principles" and that receives little support from "a government demonstrating good international citizenship." 110 A series of U.S. trade actions have been brought over time against Swedish stainless steel products,111 and the lesson, according to the author is
that good economics, international competitiveness, private ownership, and limited support from a government demonstrating good international citizenship are not enough to defend an industry against the application of antidumping or other import restricting policy. . . . [T]he Swedish government, in its compliance with OECD criteria guiding national steel policy, demonstrated better international citizenship than either the United States or the European Community.112
Mr. Finger, commenting on this case study, concludes that
Antidumping is anticompetition policy, not procompetition. . . . [T]he U.S. industry used antidumping and other unfair trade remedies to attack (Swedish) producers who had started out in the same situation as the U.S. producers but had fought their way through a disciplined, market-accommodating restructuring and downsizing to restore their profitability.113
Given such a presentation of the facts, the reader might well conclude that the U.S. antidumping law served as little more than a mechanism employed by U.S. protectionists to harass market-oriented, pro-competitive Swedish entrepreneurs.
An objective reader of the Swedish specialty steel case study might have been interested to know—but wasn't told anywhere in Finger's book—that on July 18, 1990, Avesta, one of Sweden's leading stainless steel producers and a centerpiece of his case study, was found by the EC Competition Directorate to have been a participant in the so-called Sendzimir Club or Z-mill club, a secret cartel of stainless steel sheet producers that, in the words of the Commission,
prevented, restricted and distorted normal competition in the common market by controlling production, by sharing markets and customers, and by providing the basis for concerted practices on prices. . . . [These actions] inevitably had a significant effect on conditions in the Community market.114
The details of Avesta's participation in the Sendzimir Club were set forth in detail in the EC Commission's public findings. Basically, Avesta was a signatory to a secret agreement signed in Dusseldorf on May 16, 1986, that divided up the European market for stainless steel sheet [meaning the EC and EFTA (European Free Trade Association) countries] among the participating producers, establishing a system of delivery quotas for each market, together with arrangements for fines for deliveries to any market exceeding the quotas set. A sophisticated administrative structure was set up to run the cartel, characterized by frequent meetings and periodic adjustments of the quotas.
This episode, to be sure, could be dismissed as an anomaly; it might be argued that, notwithstanding this apparent lapse by Avesta, Swedish steel producers embrace "good economic principles" most of the time. But there is more. On February 16, 1994, the EC Commission's Competition Directorate published a decision imposing sanctions on the members of a secret European cartel in the steel beams (structurals) industry, and again, Swedish steel firms were found to be deeply involved, including SSAB, the largest producer in Sweden; Ovako Profiler AB, a Swedish producer of specialty and carbon steel products; and Fundia Steel AB. The Commission's finding set forth the details of an extraordinarily complex set of arrangements involving price fixing and market division on a country-by-country basis for the EC and Scandinavia; exchange of information between producers; harmonization of charges for "extras"; and the imposition of fines on companies violating these accords. 1l5 EC Competition Commissioner Van Miert stated that
Everything that could have been infringed was infringed. It was a serious cartel involving all of the firms in the sector. It was flagrant and prolonged.116
It might be argued that European competition authorities acted against the Sendzimir and Beams cartels, thus demonstrating that national competition rules
are in fact disciplining such activity. However, the Sendzimir group was let off with small fines, reflecting, in part, the complicity of government officials in the activity concerned. The decision in the Beams case made clear that cartel activity in other product areas, and other markets, was known to the Commission but was not subject to sanctions. Public documents filed by at least one participant indicated that, notwithstanding the Commission's actions, it would take a wait-and-see attitude rather than withdrawing from other, similar arrangements in other product areas and in other international markets.117
The nexus between such anticompetitive arrangements and dumping is the same one that British critics of dumping cited at the turn of the century. Groupings such as the Sendzimir Club seek to stabilize prices in their home markets by creating an artificial constraint on supply. At the same time, they can maintain high operating rates, reduce unit costs, and enhance profitability as long as outlets exist in external markets where surpluses can be disposed of without disturbing the market order at home—that is, they are dumped.118
Although none of the activities of the Sendzimir Club or the Beams cartel found its way into Finger's analysis of antidumping and the Swedish steel industry, his omissions were not unique—they are virtually universal in critiques of antidumping policy. The existence of anticompetitive groupings that restrain supply within their own markets and dump surpluses in external markets is either not known, or if known, not reported to readers. But until such activities and their implications are fully understood and directly addressed, it is difficult to see how a meaningful public discussion of the continuing relevance of antidumping measures can go forward.
The Sendzimir and Beams cartels were operating in the European Union, which has one of the world' s most rigorous competition regimes. If such activity remains common in the EU, it is not unreasonable to suppose that it occurs with equal or greater frequency in markets where competition rules are less stringent,
to the extent they exist at all. Japan, where weak enforcement of the Antimonopoly Law has been a subject of widespread comment and criticism, is characterized by highly cartelized industries that differ little in basic structure or competitive dynamics from the German kartells of a century ago.119 And numerous newly industrializing and developing countries either have no competition laws or are only now beginning to implement them. It is surely no coincidence that the incidence of antidumping actions around the world still tends to cluster in sectors where anticompetitive arrangements are known or frequently reported to exist.
Even if the persistence of current market access barriers and restrictive private practices is acknowledged, it might be argued that because of such practices dumping remains a problem in traditional capital-intensive sectors such as cement, steel, and paper. The world economy is increasingly dominated by technology-intensive industries in which the competitive dynamics are quite different than those of a century ago. In these sectors, dumping may not be a significant problem. In fact, experience has demonstrated that dumping can be, if anything, even more destructive in its impact. It has had a devastating effect in several technology-intensive sectors, notably consumer electronics, microelectronics, and telecommunications equipment. Because at least some of these industries are as integral to national defense as steel once was, dumping can pose national security concerns as serious today as those at the turn of the century.
In high-technology manufacturing industries, research and development, and capital investment requirements are extremely high—a single semiconductor wafer fabrication facility, for example, can cost in excess of $1 billion. At the same time, in contrast to steel, product life cycles are extremely short, with the result that these investments must be entirely recovered within a very short time, typically three to four years. In this setting, intensive dumping can literally destroy an industry in a matter of months, as happened to much of the U.S. dynamic random access memory (DRAM) sector in 1984-1985. The massive losses incurred in such episodes may preclude investment in the next generation technology for that product, with the result that, from a commercial perspective, a firm is permanently driven out of a sector.
Competition in high-technology sectors is rendered more intense by "learning curve" pricing tactics, pursuant to which producers seek to maximize production volume early in a product life cycle because production costs decline in a predictable fashion based on cumulative production experience. Firms that achieve high volumes in the early stages of the product cycle achieve a cost ad-
vantage that may well prove to be commanding. However, playing this game requires massive investments accompanied by aggressive initial pricing, a high-risk approach that can culminate in market dominance, or, conversely, disastrous losses and even the destruction of the firm. "Learning curve" competition, while brutally Darwinian in its implications, is arguably consistent with market-based economics unless one or more of the competitors is operating from a protected "sanctuary" market. In that case, a company can move down the learning curve at relatively low risk through sales in the protected market. Such cost advantages and profits from the sanctuary can then be used to subsidize attacks on rival firms outside the sanctuary. The net result of diminished risk for the protected firms and exacerbated risk for other firms is familiar—disinvestment by firms adversely affected by dumping which cannot respond in kind in their rival's sanctuary.
The Japanese electronics industry, for example, is dominated by large, diversified industrial groups that have limited competition among themselves in the Japanese market in a number of product areas, with the result that prices are higher than world prices in many product lines. Profits from these "sanctuary" sectors have been used to finance aggressive entry into contested areas, selling at a loss, if necessary, for a sustained period. This dynamic has been observable in semiconductors, telecommunications equipment, and a number of other technology-intensive product areas. The high volumes that have been achieved as rivals were marginalized have eventually translated into a cost advantage and, as investment by rivals is deterred, a long-run technological advantage—in effect, the same dynamic observable in steel competition at the turn of the century.
The gradual shift in competitive advantage that took place between the British and German steel industries has close parallels in electronics. Twenty-five years ago U.S. consumer electronic firms were the world leaders in sales and technology. They sought to establish a competitive position in Japan, but were blocked by a combination of formal and informal market barriers as the Japanese government worked to foster an indigenous consumer electronics industry.120 Government restrictions included a prohibition on the establishment of local subsidiaries, foreign exchange controls, import quotas (particularly with respect to replacement and repair parts), and a 30 percent tariff.121 Although these restrictions were largely phased out in the early 1970s, an even more effective market barrier was the arrangements made by the Japanese producers themselves to restrict market penetration. Japanese consumer electronics retail outlets and service facilities were usually owned or controlled by large keiretsu-affiliated elec-
tronics firms. These firms enjoyed enormous leverage over the small retailers because of the dependency of the latter on the former for capital and because of a system of rebates given by the manufacturer to retailers for factors such as "loyalty" and "cooperation."122 Because the retail price was the same at all retail outlets, manufacturers could also reward or punish individual retailers by varying the wholesale price of merchandise. This leverage was used to prevent retailers from handling foreign merchandise.123 U.S. consumer electronics firms might have in theory set up their own dealer networks, but they were prevented from doing so by a welter of official restrictions.124
But the restrictive practices in the Japanese consumer electronics sector went far beyond such garden-variety vertical restraints. For decades a number of the major electronics producers were involved in an extraordinarily elaborate system of clandestine and semi-clandestine arrangements to regulate various aspects of market competition in consumer electronics.125 They operated a network of working groups126 that met monthly from 1964 through at least 1974 to agree on future
production and shipment levels for televisions. The JFTC did not impose sanctions for any of the anticompetitive activities engaged in by the Japanese companies,127 and its relative passivity was essential to the continued functioning of the cartel:
Regular, frequent meetings, at which manufacturers' representatives negotiated outputs and prices at the retail, wholesale, and manufacturers' levels were an important element of [the cartel's] success. . . . The story of the cartel shows that it would have been an impossible venture had the JFTC possessed more power. The members were continuously renegotiating complex, detailed agreements at numerous meetings, which did not escape notice by the JFTC.128
But it is unclear that a series of warnings and recommendations by the JFTC to the Japanese electronics firms, spanning over 30 years, has done much to curb anticompetitive behavior.129
In 1984, "evidence suggesting anticompetitive behavior in the marketing of office computers and other final products utilizing microchips" [was] uncovered by the JFTC;130 the JFTC noted the increasing "capture" of wholesalers by manufacturers, increasing producer shareholding in distributors, and transfer of management personnel between manufactures and sellers, and found that contracts signed between producers and dealers contained "restrictions regarding retail prices, sales area, retailers to whom the products could be sold, and other matters, [and] restrictions which conflict with the intent of the Antimonopoly
Act."131 In 1992 the JFTC and Ministry of International Trade and Industry (MITI) began investigations into allegations of illegal price fixing of audio-visual appliances by Matsushita, Sony, Hitachi, and Toshiba.132 In 1993 the JFTC was reportedly investigating allegations of dango (bid rigging) in connection with the sale of large-scale display screens manufactured by Sony, Matsushita, and Mitsubishi. 133 A recent German study found that 80-90 percent of the retail sales of consumer electronics products in Japan involved retailers' sales of items made by "their" domestic manufacturer, and that "the tied retailers do not usually carry directly competing products. . . . Japanese manufacturers of domestic electrical appliances have broad control over the marketing chain right down to the consumer. . . . [T]his means that the tied retailer sector, which also has service facilities, is in general not accessible to non-Japanese manufacturers and their direct marketing partner (importers)."134
Eventually, the anticompetitive practices that were endemic in the Japanese domestic electronics market spilled over into the international arena—with devastating effects. As one recent study observed, the
facts demonstrate that the seven [Japanese electronics] firms carefully coordinated their export plans; they notified one another of the intended quantity of shipments and prices, allocated U.S. customers among themselves, and cooperatively concealed a web of illegal, covert activity while charging prices low enough to suddenly and decisively gain a large share of the American market.135
In the television case, although numerous legal remedies were invoked by U.S. producers, no real answer was found either to dumping itself or to the Japanese market barriers and cartel practices that had made dumping possible. Instead, the U.S. television industry largely disappeared, foreclosing not only U.S. participation in this sector in the future, but in succeeding generations of products such as VCRs.
The problems that "downstream" British industries faced at the turn of the century as the effects of dumping gradually made them dependent on their Ger-
man competitors for key inputs has been paralleled, to a degree, in electronics. Both U.S. and European firms have repeatedly been placed in competitive difficulty as a result of their dependency on the Japanese electronics producers' group for components and tools. The most dramatic instance was the ''chip shortage" of 1987-1990. 136
Japanese semiconductor production is dominated by the same large electronics firms that comprised the television cartel described above—semiconductors are a basic component used in the end products manufactured by these firms, not only televisions but other computers, telecommunications equipment, robots, and factory automation systems. Beginning in the 1960s, these firms were organized by MITI into a series of research and development consortia for the purpose of catching up with the United States in semiconductor technology. Foreign sales in Japan were severely restricted through a ban on foreign investment, import restrictions, and local content requirements.137 Under U.S. pressure, Japan committed to eliminate these formal market barriers by 1974, and MITI undertook an urgent program of "liberalization countermeasures" designed to offset the effect of liberalization when it was implemented. Japanese semiconductor and computer firms were encouraged to form tie-ups for the production, marketing, and sales of semiconductors and computers after liberalization.138 Perhaps not surprisingly, after elimination of formal barriers to market entry in 1975, there was no increase in the foreign share of the Japanese market. U.S. firms were able to sell semiconductor products in Japan when a competing Japanese alternative did not exist, but when Japanese devices became available (often simply copies of U.S. devices), U.S. sales fell dramatically, in some cases resulting in a total loss of market.139
In the early and mid-1980s, Japanese semiconductor companies used their protected home market to pursue an aggressive trade strategy characterized by periodic episodes of dumping in the United States. 140 In the 1980s, Japanese companies repeatedly dumped DRAM and erasable programmable read-only
memory (EPROM) chips in the United States, which in the case of DRAMs, succeeded in driving virtually all U.S. firms from the market.141
Most U.S. DRAM producers withdrew from the market in mid-1985, and the Japanese DRAM producers were left with a virtual world monopoly of this strategic product. At this point, the Japanese DRAM producers began jointly curtailing their production so as to raise prices.142 In 1986 MITI announced a system of production "guideposts" (indicative production limits) designed to create a state of tight supply and higher prices.143 By 1987, the market power of the Japanese producers was so great that a "chip shortage" occurred; they exercised concerted production restraint in the face of strong demand, resulting in worldwide shortages, skyrocketing prices, and economic dislocation for foreign firms dependent on Japanese components—and enormous profits for Japanese DRAM makers.144 The Japanese firms continued to supply their own end users with DRAMs during this period, giving them a competitive advantage internationally. In addition to price manipulation through coordinated production controls, Japanese DRAM producers reportedly attempted to leverage their dominant position in the DRAM market into other markets by "tying" DRAM sales to sales of other unwanted custom chips such as application-specific integrated circuits [ASICs]. 145 Significantly, none of these experiences was replicated in EPROMs, where U.S. producers retained a much more substantial presence in the market and served as a check on the market power of the Japanese firms.
The strategic implications of U.S.-Japanese competition in electronics are not particularly difficult to discern. This competition occurred between the industries of two close allies between whom armed conflict is unlikely under any foreseeable circumstance. However, the Gulf War underscored the extent to which advanced electronic systems are likely to dominate future wars and to which U.S. forces will depend on foreign companies and systems. As other advanced electronics industries are emerging around the world, the competitive dynamics that characterized U.S.-Japanese rivalry in the 1970s and 1980s are likely to manifest themselves again, perhaps in a strategic context that is considerably less benign.
Although the character of strategic industries has shifted from sectors such as steel and dyestuffs to technology-intensive industries such as advanced electronics, new materials, and aviation, dumping poses a problem little different from that which Britain faced a century ago. The loss of strategic industries, and the resulting dependency on foreign sources for defense-related supplies, inevitably poses security risks that cannot readily be remedied once a conflict actually breaks out. Even if the foreign source of supply is a close ally, access may be
foreclosed through enemy occupation or interdiction, competing demands on the products in question, or a foreign decision to withhold supply for a variety of policy reasons.
Antidumping measures have evolved over time that have offset not only "classic" dumping (e.g., price discrimination between markets) but export sales below the cost of production.146 Commentators on antidumping frequently acknowledge that sustained below-cost export sales can be indicative of a predatory export strategy designed to drive rivals out of the market. However, it is argued that predatory schemes are rare and that assessing antidumping duties against below-cost exports penalizes certain legitimate economic activities, such as inventory clearance, "forward pricing" of new products whose average costs are initially high but decline rapidly as cumulative output increases, and export sales during recessions. In fact, predatory export strategies do appear to be extremely rare in the real world, at least as the term "predatory" is defined by contemporary economists and court decisions, and the need to protect industries against predation does not, by itself, appear to justify most of the cost-of-production antidumping duties that have been applied in recent decades.147 It is also probably true that
the scope of below-cost antidumping measures permitted under the GATT and national legislation has resulted in the application of duties in specific cases in which little policy justification exists to support the measures taken, although refinements in the GATT Antidumping Agreement limit the extent to which this can occur. It does not follow, however, that below-cost antidumping measures should be abolished or even substantially curtailed.
In a market-based economy, below-cost sales generally cannot occur indefinitely, since eventually the seller will be required to exit the market. Sustained sales at below cost are thus indicative of abnormal business behavior, and a variety of types of market distortion can give rise to sustained below-cost export sales. For example, a number of instances of below-cost dumping appear to involve the cross-subsidization of the exported product with profits generated from a domestic sanctuary market in which competition is restrained. This situation is simply a variant of Viner' s "classic" dumping and is harmful for the same reasons.
Some normal commercial practices, such as inventory clearance sales, involve sales below cost for a limited time, but under current GATT rules and most national legislation, such short-term sales do not constitute a basis for imposition of antidumping measures. Although "forward pricing" of exports based on anticipated profits is cited as a legitimate rationale for below-cost export sales, this rationale could be used to justify any below-cost export sales—a product's life cycle cannot be predicted accurately, and there is no way of knowing whether it is reasonable to expect that full costs will ever be recovered. Finally, while it is certainly true that below-cost exports occur during recessions, the vast majority of these are not subject to antidumping measures. Moreover, a serious question is presented as to
whether it is welfare-maximizing for the country of import to absorb, on a sectoral basis, somebody else's homemade recession. I dread to hear some people argue that, prima facie, this would be a good thing because it would result in low prices for consumers. If this were true, maybe we should not wait for recessions to be imported, but we should rush and ask our central banks to severely deflate so as to generate lots of low prices for consumers.148
In fact, it is unlikely that a multilateral regime that permitted unrestricted below-cost export sales—whether generated by recessions, anticompetitive syndicates, or some other factor inconsistent with market-based economies—could long endure the political pressures that would be engendered.
ANTIDUMPING AND TRADE LIBERALIZATION
For a century the concern that dumping by foreign syndicates might destroy key industries has been used by advocates of protectionism as a rationale for
restricting imports in general, whether through a high tariff wall, quantitative restrictions, tariff, or other means. A variant on this argument has been used in developing countries as ajustification for protectionist measures to defend against the feared depredations of multinational corporations. And yet despite the persistence of anticompetitive business groupings in international trade, and of dumping, the world trading order has been progressively liberalized since mid-century. The positive role played by antidumping—commonly castigated as nothing but a protectionist tool in this process—should be recognized.
The enactment and refinement of antidumping measures worldwide is almost always an element in a broader program of trade liberalization or a mechanism for defusing protectionist pressures. Typically, proponents of liberalization argue that the danger of dumping in specific sectors should be addressed through administrative measures limited to the sectors where dumping actually occurs and should not stand in the way of a more general reduction in trade barriers. The world's first antidumping legislation was enacted by Canada in 1904 by the Liberal party to neutralize domestic manufacturers' opposition to a more general reduction in import duties.149 Similarly, the first U.S. antidumping legislation, the Antidumping Act of 1916, was supported by the Wilson administration which, "while showing itself wholly sympathetic with the desire for adequate protection against unfair foreign competition, was determined that it should not be employed to build up sentiment for an upward revision of the existing tariff act."150 In the modern era, congressional political support has been sustained for the ratification of a successive road of multilateral tariff reductions, in part because the implementing legislation has incorporated refinements in U.S. antidumping procedures. A similar process is now observable in newly industrializing countries such as Mexico, Korea, and Chile, which are making greater use of antidumping measures, and strengthening their antidumping procedures, as they move to make their markets more open.
One of the most pervasive charges against antidumping policy is that it is spreading to newly industrialized countries like a sort of plague, threatening the liberal world trading order. But this reasoning implies that these countries' markets were previously open, and that a shadow is falling across this happy state of affairs as free-trading nations imitate the United States and begin to put in place antidumping regimes, a regression to a more protectionist policy. In fact, with the exception of a few special cases such as Hong Kong, virtually all newly industrializing and developing countries have been highly protectionist in the postwar era, even countries such as Taiwan and Korea, which were sometimes touted
as liberal. The market barriers in these countries were not particularly easy to identify because they consisted, typically, of opaque practices such as import licensing and prior approval requirements and the grant of import monopolies to domestic producers of the imported product.151 This situation began to change at the end of the 1980s as many newly industrialized countries and developing countries began scrapping their systems of administered protection in favor of transparent, GATT-based import regimes. The adoption of antidumping measures is part of that process of trade liberalization, and, as such, their advent should be welcomed, not condemned, provided that transparent and fair procedures are adopted. A member of the World Trade Organization Secretariat recently observed on this point that
The literature on the effects of anti-dumping duties assumes that no alternative protection would have been put in place. This assumption, however, is highly debatable. All the countries that have undergone substantial trade liberalization understand how difficult it may be to implement this policy, especially when the groups adversely affected are (politically) visible while the groups benefitting are (politically) dispersed. It is more than likely that, at least on occasion, the country of import would have let some of the steam pressuring trade reform come out in the form of additional protection. 152
It might be argued, of course, that even if some form of antidumping rules is needed as a transitional political concession to certain constituencies to facilitate the transition to liberal trade, this is a second-best solution, and that over the longer term, as the political base for open markets becomes stronger, antidumping should be phased out. The flaw in this reasoning is that if dumping really is a harmful commercial practice, foregoing its regulation will generally tend to undermine the political base for a liberal trade policy.
It is worth returning to the case of Britain in this connection. Britain did adopt antidumping regulation in the 1920s, but the rules enacted were so cumbersome as to be unusable, so that for practical purposes, dumping was permitted to continue unrestricted through the 1920s. With the onset of a world economic crisis at the end of the 1920s, Britain's 80-year consensus in support of free trade collapsed dramatically, and a wall of tariffs was erected around the Empire. Although the precise causes of this seismic shift in British economic policy have been the subject of some disagreement, it is clear that "the speed and completeness with which the remaining free trade support collapsed in 1930 can only be understood in the context of growing disillusion with trade liberalism in the late 1920s."153 Underlying this disenchantment was the persistence of an old problem—barriers in foreign markets, coupled with dumping in Britain's own. Farmers were "shocked by the intensive dumping of foreign fruits and vegetables which had destroyed markets before smallholders were able to dispose of their crops,"154 and a 1930 manifesto by British banks—long supporters of free trade—proclaimed that
Bitter experience has taught Great Britain that the hopes expressed four years ago in a plan for removal of the restrictions upon European trade have failed to be realized. The restrictions have materially increased, and the sale of surplus foreign products in the British market has steadily grown.155
The world trading system may never again confront stresses of the magnitude of those of the early 1930s, which saw an extraordinary regression into protection worldwide. But the current political consensus in support of liberal trade should not be taken for granted, if only because it ultimately rests on its members' continuing assessment of where their self-interest lies. Antidumping measures are a safety mechanism not only for defusing protectionist pressures, but for the wholly legitimate purpose of limiting a harmful commercial practice that, left unchecked, could undermine support for the current system.
Antidumping as an Interface Mechanism
Antidumping measures have become controversial, in significant part, because they have been assigned, by default, an impossible task—to reconcile the economic and strategic contradictions that arise out of the sharp divergences that exist between national markets with respect to competition policy. The fact that some markets are open and others are highly cartelized gives rise to distortions in trade and economic dislocations for which national antitrust policies and the multilateral system have no apparent answers. Antidumping measures are invoked by beleaguered industries because nothing else works, including improvements in their own efficiency and productivity. The application of duties at the border in a given instance may prevent the destruction of an industry by dumping, but it does not resolve the market distortions that gave rise to dumping in the first place.
Virtually all of the extant literature on antidumping measures emphasizes the problems that such measures allegedly create, rather than on whether or not they are actually effective in addressing the problem at which they are directed—dumping. Thus, specific instances are raised in which antidumping measures are said to have been applied to inappropriate situations, or in a way that is unnecessarily burdensome or that results in margins of dumping that are too high. In some cases the criticism is valid and provides the basis for future reforms both at the national and the multilateral level. At the same time, however, antidumping measures do not always prevent serious injury to affected industries.156 Margins are not always high enough to fully offset the injurious effects of dumping, and antidumping orders can be circumvented through a wide variety of commercial tactics. In high-technology industries the sector may be largely destroyed even before preliminary relief is available. Antidumping actions are burdensome to petitioners as well as respondents and the cost of such proceedings has mounted as the information required of petitioners has increased. 157
If dumping itself remains a "problem in international trade," then true "reform" of antidumping policy does not simply entail weakening or eliminating national antidumping laws, but the shaping of those laws to rectify, to the fullest extent possible, the problem of dumping itself. Antidumping measures cannot, by themselves, open foreign markets or break up cartels, but they can form one element in a broader program aimed at such objectives. Antidumping should not disrupt exporters that are not engaging in anticompetitive practices (as now occurs), but it should offer the most efficient and complete relief possible to industries genuinely injured by dumping. Such reform is unlikely to emerge from a debate cast in the simplistic free-trade versus protectionism terms which charac-
terized the British policy debates at the turn of the century. Effective pragmatic reform requires a far more informed and dispassionate examination of dumping itself and the closed markets and cartels that foster it. Only such a comprehensive approach, rooted in the realities of commercial practice, will make possible reforms necessary to enable the multilateral trading system to adjust to the challenges of the next century.