Other candidates come to mind when contemplating the world’s worst international trade scofflaw, but the United States makes a strong case for itself. A recent Commerce Department determination that foreign companies like Samsung, LG, and Electrolux engaged in “targeted dumping” by reducing prices on their washing machines for Black Friday sales confirms that the United States is actively seeking that ignominious distinction.
U.S. policies have been the subject of more World Trade Organization disputes (119, followed by the EU with 73, then China with 30) and have been found to violate WTO rules more frequently than any other government’s policies. No government is more likely to be out of compliance with a final WTO Dispute Settlement Body (DSB) ruling — or for a longer period — than the U.S. government. To this day, the United States remains out of compliance in cases involving U.S. subsidies to cotton farmers, restrictions on Antigua’s provision of gambling services, country of origin labeling requirements on meat products, the so-called Byrd Amendment, a variety of antidumping measures, and several other issues, some of which were adjudicated more than a decade ago. In some of these cases, U.S. trade partners have either retaliated, or been authorized to retaliate, against U.S. exporters or asset holders, yet the non-compliance continues as though the United States considers itself above the rules.
Despite all the official high-minded rhetoric about the pitfalls of protectionism and the importance of minding the trade rules, the U.S. government is a serial transgressor. Nowhere is this tendency to break the rules more prevalent than it is with respect to the Commerce Department’s administration of the antidumping law. Nearly 38 percent (45 of 119) of the WTO cases in which U.S. policies have been challenged concern U.S. violations of the WTO Antidumping Agreement. The epidemic of targeted dumping findings — such as the one concerning washing machines, if the U.S. International Trade Commission renders an affirmative injury finding next week — will more than likely produce new WTO cases.
The antidumping law is purported to exist to protect American companies and their workers from the effects of foreign competitors selling their products in the United States at “unfairly low” prices. Why price competition — encouraged as it is between domestic rivals — suddenly becomes unfair or worth thwarting when foreigners are offering the lower prices is a question without a satisfactory answer. The public is under the false impression that trade is a contest between Team America and the foreign team, and that they should cheer when our government erects barriers to defend us against foreign commercial success. The persistence of that mindset shields U.S. antidumping policy from the scrutiny it deserves.
Instead, the law has become a mechanism through which domestic companies — with the assistance of creative lawyers and a captured government agency that is committed to keeping America safe from imports — can saddle their competition (both foreign and domestic) with higher costs, control supply, increase their own prices, and reap higher profits. Nevermind that this “success” comes at the expense of consumers and firms in downstream industries that require access to the restricted product.
While President Obama urges U.S. companies to become more competitive at home and abroad, the U.S. antidumping law kneecaps these very same enterprises by making their industrial inputs and intermediate goods scarce and more expensive — as described and documented in this paper. It also inspires retaliatory protectionism abroad.
In general terms, dumping is defined as the sale of a product in a foreign market at a lower price than the price obtained by the same producer in his home market. Dumping is measured by comparing a foreign producer’s U.S. and home market prices over a specific period of time. For each comparison, the difference between the U.S. price and the home market price is considered the unit margin of dumping. A positive dumping margin results when the U.S. price is lower than the home market price and a negative dumping margin is the result when the U.S. price exceeds the home market price. The antidumping duty ultimately imposed is, in theory, equal to the weighted average dumping margin calculated for all U.S. sales expressed as a percentage of U.S. sales value.
Under the WTO Antidumping Agreement (ADA), governments are permitted to have antidumping laws and to apply antidumping duties to redress dumping that is found to be a cause of material injury to the domestic industry producing the same or similar products. Although the ADA is deferential to national governments when it comes to the details of implementing their antidumping laws, it does articulate certain minimum standards intended to limit the scope for abuse, such as reaching affirmative findings of dumping when no dumping has occurred or manufacturing punitively high antidumping duty rates, for example.
But this seemingly mechanical exercise of comparing prices and calculating margins is rife with subjective interference and methodological sleights of hand. Under the U.S. antidumping law, the Commerce Department maintains considerable discretion when it comes to determining the existence and measuring the magnitude of dumping. Which sales should be included in calculating average prices? What product models should be collapsed together and treated as a single model for purposes of calculating average prices? What expenses should be subtracted from gross prices before net prices are compared between markets? Is there evidence of targeted dumping?
Those are just a few of the many kinds of consequential, results-changing decisions the Commerce Department renders during the course of an antidumping proceeding. That broad discretion has been abused over the years, as confirmed by the hundreds of U.S. court rulings that have found the Commerce Department acting illegally or otherwise beyond it authority.
Just as U.S. antidumping administration has so frequently run afoul of the U.S. law, it has also been found on dozens of occasions to violate U.S. obligations under the WTO Antidumping Agreement. In 12 of the 45 cases in which ADA violations were alleged by U.S. trade partners, the methodological trick known as “zeroing” was at least one of the subjects of controversy. Ultimately, this issue is the motivation behind the Commerce Department’s shenanigans in the case involving imported washing machines, targeted dumping, and Black Friday sales.
“Despite all the official high-minded rhetoric about the pitfalls of protectionism and the importance of minding the trade rules, the U.S. government is a serial transgressor.”
To appreciate the affront to fairness and mathematical integrity that zeroing represents, recall that antidumping duty rates are determined by calculating a weighted average dumping margin, which derives from individual dumping margins calculated for multiple comparisons. Some of those comparisons produce positive dumping margins (whenever the U.S. price is lower than the home market price) and some yield negative dumping margins (whenever the U.S. price is higher than the home market price). Zeroing refers to the practice of assigning those negative dumping margins a value of “ 0″ before calculating the weighted average dumping margin, which has the effect of increasing the weighted average dumping margin and the applied antidumping duty rate.
Analysis from the Cato Institute found that zeroing, in a sample of 18 actual antidumping case records reviewed, artificially inflated antidumping duties by 44 percent. Here’s how it works in practice. Let’s say a foreign widget producer makes two sales in the United States. He sells one widget at a net price of $1.00 and the other at a net price of $3.00. In his home market, he sells the same widgets for $2.00. Is he dumping? There is a dumping margin of $1.00 on the first sale ($2.00 — $1.00 = $1.00) and a dumping margin of minus (-)$1.00 on the second sale ($2.00 — $3.00 = -$1.00). The weighted average dumping margin is 0 and thus there is no dumping. But when the Commerce Department engages in zeroing, it treats the -$1.00 dumping margin as equal to $0.00, denying its impact on the overall weighted average dumping margin. By zeroing, the Commerce Department would find a total dumping margin of $1.00 and express it over the total value of all U.S. sales ($4.00), which would yield an antidumping duty rate of 25 percent. Magic!
Zeroing has been found by the WTO Appellate Body — on numerous occasions and under multiple comparison methodologies — to violate Article 2.4 of the ADA because it precludes consideration of the impact of all export sales in determining the existence and magnitude of dumping, and thus precludes a fair comparison. After many years of appeals and foot-dragging, the United States finally changed its policy to comply, grudgingly, with the jurisprudence that has shut the door on zeroing under all circumstances, using all comparison methodologies…with the possibility of one tiny exception. (This article describes the evolution of the zeroing jurisprudence and provides a detailed analysis of this “tiny exception.”)
Article 2.4.2 of the Antidumping Agreement states that the existence of margins of dumping during the investigation phase “shall normally be established” by comparing prices on an average-to-average or transaction-to-transaction basis, but may be conducted on a home market average-to-individual export transactions basis:
[I]f the authorities find a pattern of export prices which differ significantly among different purchasers, regions or time periods, and if an explanation is provided as to why such differences cannot be taken into account appropriately by the use of a weighted average-to-weighted average or transaction-to-transaction comparsion.
The justification for the exception under Article 2.4.2 is, presumably, that that comparison methodology allows the authorities to home in on targeted dumping, which may be revealed by patterns of price differences among purchasers, regions or time periods. Some have argued that zeroing must be permissible under the exceptional method because otherwise its prohibition would yield results identical to those obtained under the weighted average-to-weighted average method. This “mathematical equivalence” argument, as it is called, holds that there would be no practical reason for the exception to exist if zeroing were precluded, and that would then violate the rules of effective treaty interpretation by rendering the exception inutile.
And, finally, this takes us back to Commerce Department’s determination that foreign washing machine manufacturers are engaging in targeted dumping. The Commerce Department is attempting to operationalize the mathematical equivalence argument by making this tiny, rarely-ever-used exception the new rule so that it has license to engage in zeroing and inflate antidumping duty rates on behalf of certain domestic producers. But it is carelessly and defiantly overplaying its hand.
For starters, the standards that Commerce is trying to establish for finding a “pattern of export prices which differ significantly among different purchasers, regions or time periods” have no statistical rigor, no theoretical underpinnings, and no relation whatsoever to unmasking, isolating, or measuring targeted dumping. Two arbitrary benchmarks must be met to find targeted dumping. First, at least 33 percent by volume of the sales to an allegedly targeted group must be at prices that are less than one standard deviation below the average price to all other groups. Why 33 percent? Why one standard deviation? Neither figure is supported by statistics literature or law.
Second, at least 5 percent of the volume of sales to the allegedly targeted group must be at prices that are lower than the average price of the lowest-priced non-targeted group by a margin (a gap) that is greater than the average gap between the average prices of all the non-targeted groups. The idea here is that the allegedly targeted group should be uniquely identifiable as having received significantly lower prices. The problem again, though, is the arbitrary nature of the 5 percent threshold. Seems way too low anyway.
With those benchmarks satisfied, the Commerce Department considers itself free to measure dumping by using the weighted average-to-transaction methodology for all sales, and to engage in zeroing for all sales.
Beyond these obvious shortcomings is the fact that if the application of zeroing to all sales to inflate dumping margins is automatically triggered by the idenfication of any customer or any region (however infinitesimally small petitioners wish to define it) or any time period (ranging anywhere from one day to one year) that meets the benchmarks for a targeted dumping analysis without need of any plausible explanation as to why the particular group would be targeted, then every single petitioner in every subsequent case will pour over the numbers and run their own statistical analyses to identify just one group, any group, however imaginatively defined, that satisfies those benchmarks. Black Friday sales are an obvious candidate for a targeted group.
This whole enterprise is an absurdity that other governments should and will find offensive.
In addition to the shortcomings identified above is the fact that the Commerce Department’s statistical tests are not even doing what the Commerce Department thinks they’re doing. By comparing average prices instead of actual prices, the standard deviations it uses as benchmarks are significantly tighter because the range of prices, which produce the standard deviations, has been squeezed tighter by averaging them first. So, beyond all of the theoretical objections are these practical misapplications that, at the very least, should be remanded back to Commerce by the courts.
The Commerce Department’s sloppiness appears to be the product of a brute force effort to clear any annoying obstacles in its drive to recuscitate zeroing as the default practice of the U.S. antidumping authorities.
Finally, even if the exceptional method is permitted in any given case, that doesn’t grant license for zeroing as Commerce conceives of it. There should be no question that that approach will run afoul of the Appellate Body’s interpretation of what is permitted under the exceptional method. The AB has ruled on numerous occasions that fealty to Article 2.4.2 cannot come at the expense of fealty to the fair comparison language of Article 2.4, and zeroing all sales clearly commits that error.
Unmasking and measuring the extent of targeted dumping does not necessitate resort to the exceptional method for all sales. In fact, measuring targeted dumping can be achieved by resorting to the exceptional method and allowing zeroing for only the targeted groups, while using the normal, average-to-average method without zeroing for all other comparisons. This invalidates the “mathematical equivalence” argument. But the Commerce Department will no doubt have to revise its targeted dumping tests first.
The United States is definitely one of the world’s biggest trade scofflaws and the antidumping regime remains fertile ground for more mischief.