The WTO Ruling on U.S. Country of Origin Labeling (“COOL”)

Introduction

The WTO dispute on country of origin labeling (“COOL”) requirements for imported livestock is the latest in a series of cases dealing with the Agreement on Technical Barriers to Trade (“TBT Agreement”).[1] The Appellate Body Report was circulated on June 29, 2012.[2] This case pitted U.S. cattlemen against large packers and food processors and raised questions about the significance of country of origin labeling when it comes to integrated and international supply chains. This Insight provides an overview of the key issues addressed by the Panel and the Appellate Body (“AB”).

Factual Background

In this dispute, Canada and Mexico[3] challenged the legality of various U.S. laws[4]—the COOL measure—that require muscle cut meat from imported and domestic livestock (cattle and hogs) to be sold at retail with one of the following labels:

Category A
Label A – “Product of the United States” for meat derived from animals exclusively born, raised, and slaughtered in the United States.

Category B
Label B – “Product of the United States, product of country X” for meat derived from animals not exclusively born, raised, and slaughtered in the United States. This could include livestock born in Canada and raised and slaughtered in the United States. The label could also be affixed to the commingling on the same production day of categories A&B, A&C, B&C, and A, B&C.

Category C
Label C – “Product of country X, product of the United States” for meat derived from animals imported into the United States for immediate slaughter. This label could also be affixed to meat from the commingling on the same production day of meat from categories A&B, A&C, B&C, and A, B&C .

Category D
Label D – “Product of country X” for one hundred percent imported foreign meat.

The impacts of the COOL measure on the U.S. agriculture industry have been hotly debated. On one side of the debate, meat packers, processers, and retailers argue that the integrated nature of the North American livestock market makes tracing the origin of livestock costly, and that no health differences exist between U.S., Canadian, and Mexican beef because imported meat must satisfy the same U.S. health and safety standards as U.S. meat. [5] Supporters of COOL, such as U.S. cow and calf producers, argue that the incidence of mad cow disease (“BSE”) in Canada means consumers have a health interest in knowing the origin of their meat[6] and are willing to pay a premium for this information.[7]

The Panel Decision

The Panel found that the COOL measure is a technical regulation as defined in Annex 1.1 of the TBT Agreement.[8] Canada and Mexico also claimed that a letter from U.S. Secretary of Agriculture, Thomas Vilsack, to the agriculture industry addressing how companies could implement COOL was a technical regulation. The Panel rejected those claims because compliance with the letter was not mandatory—a key requirement under the TBT Agreement.[9] Nevertheless, the Panel found that the Vilsack letter breaches Article X:3(a) of the General Agreement on Tariffs and Trade 1994 (“GATT”) because it does not meet the requisite standards of transparency and procedural fairness.[10] These findings were not appealed.

The Panel found that the COOL measure breaches the national treatment commitment in TBT Article 2.1 by treating imported livestock less favorably than like domestic livestock. The Panel observed that under GATT Article III:4, products distinguished solely on the basis of origin are deemed “like products.” As the COOL measure distinguishes solely on the basis of country of origin, U.S. and Mexican livestock, and U.S. and Canadian livestock, are like products.[11] This finding was not appealed. The Panel then assessed whether the COOL measure “modifies conditions of competition in the relevant market to the detriment of imported goods.”[12] The Panel found that the effect of the COOL measure is to require “an unbroken chain of reliable country of origin information with regard to every animal and muscle cut,” necessitating segregation of livestock by origin.[13] The costs of segregation create an incentive to process U.S. meat only, thus modifying the conditions of competition to the detriment of imported Canadian and Mexican livestock.[14]

The Panel also found that the COOL measure breaches the requirement in TBT Article 2.2 that technical regulations not be more trade-restrictive than necessary to fulfill a legitimate objective, taking account of the risks non-fulfillment would create. The Panel found that the use of labels B and C interchangeably for commingled livestock would make consumers unsure about the origin of the meat.[15] As a result, the COOL measure does not fulfill the United States’ legitimate objective of providing consumers with information about the origin of meat.[16]

The Appellate Body Decision

The United States, Canada, and Mexico appealed different aspects of the Panel’s decision. The AB upheld the Panel decision that the COOL measure violates the TBT Article 2.1 national treatment commitment (but for different legal reasons) and reversed the Panel’s finding that the measure breaches TBT Article 2.2.

TBT Article 2.1

The AB reaffirmed that GATT Article III.4 jurisprudence provides “relevant guidance” for interpreting the national treatment commitment in TBT Article 2.1.[17] The AB agreed with the Panel that the COOL measure modifies the conditions of competition in the market to the detriment of livestock imported from Mexico and Canada. In order for retailers to have the information to comply with the COOL measure and affix an appropriate label, upstream meat producers segregate livestock according to origin to enable record-keeping regarding livestock origin. Because imported livestock constitute a small percentage of the overall U.S. market, the least expensive way of complying with the COOL measure is to avoid segregation by relying exclusively on U.S. livestock.[18] And while the COOL measure does not require segregation by the meat producers, creating “incentives for private actors systematically to make choices in ways that benefit domestic products to the detriment of like imported products” constitutes less favorable treatment in breach of TBT Article 2.1.[19]

The AB criticized the Panel for failing to undertake the next step by determining whether “the detrimental impact stems exclusively from a legitimate regulatory distinction, or whether the COOL measure lacks even-handedness.”[20] According to the AB, an absence of even-handedness arises when the measure is designed or applied in a manner that constitutes a means of arbitrary or unjustifiable discrimination.[21] A Panel must identify the relevant regulatory distinctions drawn by the measure.

The COOL measure distinguishes between the three stages of production—where the livestock was born, raised, or slaughtered—and between the four different labels.[22] According to the AB, these distinctions do not provide consumers with information as to the origin of the livestock that is “commensurate” with the type of origin information upstream producers are required to gather about where livestock was born, raised, or slaughtered.[23] For instance, due to the interchangeability of labels B and C for commingled livestock, consumers cannot identify where the livestock was born, raised, or slaughtered. Additionally, the information conveyed by using two country names on labels B and C is confusing. This led the AB to conclude that no “rational basis” exists for the large amount of information that upstream producers are required to collect in contrast to the small amount of information conveyed to consumers via the labels.[24] As a result, the regulatory distinction cannot be explained by the need to provide information to consumers. The AB therefore found that the detrimental impact of the COOL measure on Canadian and Mexican livestock imports does not arise from legitimate regulatory distinctions.

TBT Article 2.2

To establish whether a measure is more trade-restrictive than necessary, a Panel must assess: 1) the degree of contribution of the measure to the legitimate objective; 2) the trade restrictiveness of the measure; and 3) the nature of the risks and gravity of the consequences arising from non-fulfillment of the measure’s objectives.[25]

The AB upheld the Panel’s finding that the objective of the COOL measure is to provide consumers with information as to origin. However, the AB criticized the Panel for its willingness to base this conclusion on the U.S. classification of the measure, emphasizing that a measure’s objective should instead be determined by considering its “design, architecture, structure, legislative history, and evidence relating to its operation.”[26]

The U.S. objective of providing consumers with information as to origin is not in the list of legitimate objectives of TBT Article 2.2. However, the AB reaffirmed that this list is not closed: an objective “linked or related to” a legitimate objective listed in TBT Article 2.2 is more likely to be legitimate.[27] The preamble to the TBT Agreement and to other WTO Agreements may also inform whether an objective is legitimate.[28] The AB observed that providing consumers with information about origin could help prevent deceptive practices—an objective reflected in TBT Article 2.2 and GATT Article XX(d). The AB ultimately accepted the Panel’s conclusion that the U.S. objective is legitimate on the grounds that the complainants had failed to satisfy their burden of proof and establish that this was not the case.[29]

The final issue for the AB was whether the COOL measure fulfills the legitimate objective of the United States. The AB reaffirmed its finding in US—Tuna II that this question relates to the degree the measure contributes to realizing the legitimate objective, and that a measure need not reach any minimum threshold in order to fulfill a legitimate objective.[30] Moreover, this determination is to be based on the design, structure, operation, and application of the measure.[31] The AB noted that the Panel had found that even labels B and C—the most confusing of the labels—provide some information as to origin and information in addition to that previously available. On this basis, the AB overturned the Panel finding that the COOL measure does not fulfill the legitimate objective of providing information as to origin.[32] In the absence of sufficient evidence to determine whether a less trade-restrictive alternative was reasonably available, the AB also overturned the Panel finding that the COOL measure is inconsistent with TBT Article 2.2.

Conclusion

This case follows recent AB decisions on the TBT Agreement in US—Clove Cigarettes and US—Tuna II and further clarifies what constitutes a legitimate regulatory distinction under TBT Article 2.1. In particular, the AB finding that the requirement of even-handedness is not met when the measure is designed or applied in a manner that constitutes a means of arbitrary or unjustifiable discrimination reflects the exceptions provision in the GATT Article XX chapeau. Indeed, the absence of an equivalent exceptions provision in the TBT Agreement is what drives the AB’s approach in this case.[33] Reliance by the AB in future cases on GATT Article XX jurisprudence to inform the meaning of arbitrary and unjustifiable discrimination is now likely.

This decision also has important implications for the design by members of labeling schemes. While the AB did not directly address how much of a contribution the COOL label needs to make to the United States’ legitimate objectives, its reliance on the Panel’s finding that it is enough for the COOL label to provide more information than was previously available sets a low bar. Moreover, the AB did not impose any particular standards in terms of the quantity or clarity of information the label needs to convey. This leaves members with significant flexibility when designing their labeling schemes.

About the Author:
Joshua Meltzer is a Fellow at the Brookings Institution in Washington D.C. and an Adjunct Professor at Georgetown Law and John Hopkins School of Advanced International Studies.

Endnotes: