1. Introduction
China's trade dispute with the European Union (EU) over modifications of poultry tariff concessions allowed a WTO panel to clarify member obligations concerning both the allocation of tariff-rate quotas (TRQs) (Article XIII GATT) and the renegotiations of schedules (Article XXVIII GATT). The modification in question involved replacing several tariff lines with TRQs that were allocated largely to Brazil and Thailand, leaving China facing a higher out-quota rate for the bulk of its poultry meat exports. In such circumstances, WTO members can be consulted in renegotiations that allow them to withdraw substantially equivalent concessions from the importing nation, but here the EU deemed it was not necessary to consult with China, because China did not have a ‘substantial interest’ (per Article XXVIII GATT) in supplying poultry meat. The exclusion of China from renegotiations was thus central to the dispute.
While the Panel rejected most of China's claims, including the key claims for inclusion in the renegotiations of schedules under Article XXVIII, the Panel agreed with China that its increased ability to export poultry products to the EU was a ‘special factor’ that should have been considered in the EU's allocation of TRQs under Article XIII:2(d) GATT. China's exports had in fact increased substantially for two duck-meat tariff lines after certain SPS measures expired. On the other hand, China had much lower shares of the EU's imports for the other poultry meat tariff lines (by 2011, the shares were about 5% for two lines, and zero for the rest), so the Panel found no violation in the EU's TRQ allocation for these tariff lines.
In rejecting most of China's claims, the Panel took a limited view of the EU's obligations, due to an overly textual analysis of individual GATT articles, while undervaluing obligations according to the broader object and purpose of the GATT. Renegotiations in the GATT give nations flexibility to modify obligations, while permitting select members with a substantial interest in the modifications to withdraw equivalent concessions. The limitation on who is included in renegotiations is necessary to facilitate renegotiations, but the contract is explicit that this limitation is not subject to a precise definition (per paragraph 7 of the interpretive note of Article XXVIII:1 GATT). The Panel though rejected China's claims on the basis that the text did not provide any specific obligations for the EU in line with China's claims, while failing to consider more broadly China's claim of substantial interest. Similarly, narrow interpretation led the Panel to find no further violation in the TRQ allocation under Article XIII GATT.
Given the gap left in defining substantial interest, economic analysis can then serve a role in interpreting the GATT's overall object and purpose. Recognition of the economic purpose of renegotiations supports a more inclusive criterion for who participates in renegotiations involving tariff-to-TRQ modifications. The Panel's reasoning largely rules out violations for any importer who creates a TRQ by setting a higher out-quota rate for any country with import share less than 10% based on a specific reference period. This reasoning comes from a narrow and rigid reading of obligations from the interpretive note of Article XXVIII:1 GATT. However, the interpretive note offers leeway to define substantial interest so that renegotiations serve the WTO's broader purpose of facilitating rules-based gradual liberalization. A modification from tariff-to-TRQ for all countries with less than a 10% share is effectively a discriminatory tariff increase that, under the Panel's ruling, cannot be deterred by members withdrawing concessions, so there will henceforth be strong incentives to undertake such modifications. TRQs are dynamically inferior to tariffs, because optimal trade quantities can evolve over time, so TRQs’ discriminatory effects can cause trade diversion, as more suppliers that are efficient are unable to export. A relatively permissive view toward the modification of tariffs to rigid TRQs is unequivocally a wrench in the WTO system.
The Panel's ruling is of particular importance due to both the limited WTO jurisprudence on renegotiations and TRQs, as well as the increasing potential importance of both renegotiations of schedules and allocation of TRQs under the current wave of rebalancing demands coming from some WTO members. For example, TRQ renegotiations have already become a thorny issue for WTO compliance in the context of the impending exit of the United Kingdom from the EU, as the proposed division of TRQs between the United Kingdom and the EU would reduce export flexibility of other members (Downes, Reference Downes2017). By taking a narrow view of nations’ obligations to other members in TRQ renegotiations, the Panel's ruling limits how such renegotiations can serve the broader purpose of the WTO.
2. Summary of the dispute and Panel decision
In April 2015, China requested consultations with the European Union (EU) regarding measures to modify the European Union tariff concessions on certain poultry meat products (EU–Poultry Meat (China)),Footnote 1 claiming these measures were inconsistent with Articles I, II, XIII, and XXVIII GATT.
The EU initiated two distinct tariff renegotiation rounds under Article XXVIII:5 of the GATT 1994 (which allows Members to reserve their right to renegotiate and eventually exercise this right at a later date). The first round was initiated in 2006, and covered three tariff concessions (‘First Modification Package’); the second round was initiated in 2009, and covered seven other tariff concessions (‘Second Modification Package’). In both rounds of negotiations, having notified its intention to modify its tariff concessions under Article XXVIII GATT, the EU determined that Brazil and Thailand were the only WTO Members that held a ‘principal’ or ‘substantial’ supplying interest in the tariff concessions at issue. The EU based such determination on the share of imports into the EU that different Members had held over the three years preceding the initiation of each of the two negotiation rounds (i.e. 2003–2005 for the First Modification Package, and 2006–2008 for the Second Modification Package). Following the negotiations and bilateral agreements reached with Brazil and Thailand, the EU replaced its unlimited tariff concession on each of the poultry products at issue with a TRQ. The EU allocated the majority of each of those TRQs to Brazil and/or Thailand, because these two Members accounted for the majority of imports of these products into the EU in the years preceding the initiation of the negotiation rounds in 2006 and 2009. The total volume of each TRQ, and the allocation of each TRQ among supplying countries, was based on the import levels into the EU over the same period (i.e. 2003–2005 for the First Modification Package, and 2006–2008 for the Second Modification Package).
From January 2002 through to July 2008, imports of poultry products from China into the EU were prohibited as a consequence of several SPS measures (the WTO-consistency of which was not at issue in this dispute). Following a relaxation of the SPS measures in July 2008, and while the negotiations under the Second Modification Package were ongoing, imports of poultry products from China under two of the seven tariff lines at issue in the Second Modification Package increased significantly over the period 2009–2011. By the time that the EU concluded its negotiations under the Second Modification Package in late 2011, China accounted for more than 50% of imports into the EU under those two tariff lines.
China's claims principally revolved around (a) the procedure for renegotiating tariff protection under Article XXVIII GATT and (b) the requirements under Article XIII GATT for the application of the new import restrictions (i.e., TRQs) on the products at issue.
With regard to the former, the Panel rejected China's claims that the renegotiation rounds were inconsistent with both Article XXVIII:1 GATT and Article XXVIII:2 GATT read in conjunction with paragraph 6 of the Understanding on Article XXVIII GATT. First, the Panel concluded that China had not demonstrated that the EU acted inconsistently with Article XXVIII:1 by determining which Members held a principal or substantial supplying interest on the basis of ‘actual’ import levels over the three-year period preceding the notifications to modify the tariff concessions as opposed to basing its determination on what Members’ share ‘would have been’ in the absence of the SPS measures that restricted Chinese poultry imports. The Panel also rejected China's argument that an importing Member is legally required to re-appraise those Members that hold a principal or substantial supplying interest, to reflect changes in import shares that took place during the negotiations.
Second, the Panel also rejected China's arguments that Article XXVIII:2 and paragraph 6 of the Understanding required the EU to (a) calculate the total amount of TRQs for both renegotiation rounds on the basis of an estimate of what import levels would have been in the absence of the SPS measures; (b) calculate the total amount of the TRQs for the second round on the basis of import levels over the 2009–2011 (rather than the 2006–2008) period; and (c) account for poultry imports into Romania, Bulgaria, and Croatia in the years before they acceded to the EU (in the context of the second round of renegotiations).Footnote 2 The Panel found that the EU was not required to do any of the above (inter alia because Article XXVIII:2 and paragraph 6 do not apply to the allocation of TRQ shares among supplying countries).Footnote 3 The Panel thus concluded that China had not demonstrated that the EU had acted inconsistently with Article XXVIII:2 and paragraph 6 of the Understanding by failing to maintain a general level of reciprocal and mutually advantageous concessions not less favourable to trade than those exiting prior to the modification.Footnote 4
With respect to the requirements under Article XIII GATT for the application of import restrictions, the Panel accepted some of the claims brought by China and rejected others.
With regard to the requirements under Article XIII:2(d), the Panel rejected China's claim that the EU had acted inconsistently with Article XIII:2(d) by determining which countries had a substantial supplying interest on the basis of their actual share of imports during the reference period and not based on an estimate of what import shares would have been in the absence of the SPS measures restricting poultry imports from China.Footnote 5 However, with respect to two of the ten TRQs at issue, the Panel did find that the EU's allocation of TRQ shares among supplying countries was inconsistent with the requirements of Article XIII:2(d) because China's increased ability to export poultry products to the EU following the relaxation of the SPS measures in July 2008 was a ‘special factor’ that had to be taken into account by the EU when determining which countries had a ‘substantial interest’ in supplying the products concerned. For this reason, the Panel found that the EU acted inconsistently with Article XIII:2(d) by not recognizing China as a Member holding a ‘substantial interest in supplying the products’ under two of the ten tariff lines and by failing to seek agreement with China on the allocation of the TRQs for those particular tariff lines.Footnote 6
With regard to the requirements under the Chapeau of Article XIII:2(d), the Panel first rejected China's claim because China had not demonstrated that the EU acted inconsistently with the chapeau of Article XIII:2 by determining TRQ shares allocated to ‘all others’ countries (that were not recognized as substantial suppliers – including China – on the basis of their actual share of imports into the EU) based on the actual share of imports into the EU, as opposed to an estimate of import shares in the absence of the SPS measures.Footnote 7 Second, the Panel found that the EU did act inconsistently with the Chapeau of Article XIII:2 by not considering China's increased export ability as a ‘special factor’ for purposes of determining the TRQ shares to be allocated to the category of ‘all other’ countries that were not recognized as substantial suppliers pursuant to Article XIII.Footnote 8 Third, the Panel concluded that China had not demonstrated that the EU acted inconsistently with the Chapeau by not allocating to ‘all other’ countries at least 10% of the share of the TRQs under both modification packages.Footnote 9
With regard to China's claim that the EU had violated Article XIII:4 GATT by refusing to enter into ‘meaningful consultations’ with China with a view to readjusting the allocation of TRQs, the Panel made the following findings. First, the right to request consultations under Article XIII:4 GATT is also available to Members holding a substantial supplying interest where the allocation of TRQs is agreed pursuant to Article XIII:2(d).Footnote 10 Second, the determination of which Members hold a substantial supplying interest for purposes of Article XIII:4 cannot be based solely on import shares held during the reference period initially used to determine which Members held a substantial supplying interest under Article XIII:2 without taking into account changes in market shares that occurred following the initial TRQ share allocation.Footnote 11 And accordingly, the Panel concluded that China held a substantial supplying interest in supplying two of the ten products at issue at the time of its request for consultation in December 2013.Footnote 12 Third, while the Panel found that Article XIII:4 can be interpreted as establishing a legal obligation on the importing Member to reallocate TRQ shares among supplying countries (to reflect changes in the import shares held by different countries), there is no obligation to reallocate the TRQ shares within any particular time frame, and at least not in the years immediately following the initial TRQ allocation.Footnote 13 Accordingly, the EU did not violate Article XIII:4 when it refused to reallocate the TRQ allocations (arising from the second modification package) in May 2014.Footnote 14 Lastly, the Panel concluded that China had failed to discharge its burden to demonstrate that the EU had indeed refused to consider the need for an adjustment of the TRQ shares or the reference period or a reappraisal of special factors.Footnote 15
China also made additional claims under both Article I:1 concerning TRQs violating MFN obligations and Article II:1 concerning a violation of the EU's bound rates. The Article I claim was rejected on the basis that TRQs are governed under Article XIII. The Article II:1 claim arose because China claimed the EU's modifications needed WTO certification before implementation, but the Panel found that no such obligation exists. Given the straightforwardness of these rulings, the remainder of this paper focuses instead on TRQ allocation and the renegotiation of schedules.
3. Legal analysis
What we attempt to do in this section is to investigate whether, in its analysis of the various legal provisions applicable to the renegotiation of tariff commitments and allocation and reallocation of TRQs, there were opportunities for the Panel to adopt an alternative (perhaps bolder) approach that would have better aligned the legal disciplines at issue with the broader object and purpose of the GATT, facilitating rules-based gradual liberalization.
In suggesting that alternative readings were indeed available, we highlight the Panel's inconsistent over-emphasis on the objective of guaranteeing efficiency of renegotiation under Article XXVIII (over the need to guarantee fairness of the renegotiation) and on textual interpretation (including on the lack of specific guidance in the text of the relevant provision at issue).
3.1 Determining a Member with principal or substantial supplying interest: stairway to hell?
Article XXVIII GATT sets out the procedures according to which any Member is allowed to renegotiate (for whatever reason) its tariff protection. This is part of the general flexibility embedded throughout the GATT and geared in this specific context to encourage Members to agree to ‘generous’ tariff reduction in the first place (safe in the knowledge that such reduction could be reversed in the future as long as appropriate compensation is given).Footnote 16 As noted by Mavroidis (Reference Mavroidis2016), Article XXVIII:1 thus attempts to find a balance between, on the one hand, the need to compensate those Members negatively affected by the tariff renegotiation (fairness) and, on the other hand, the need to ensure the speedy completion of the renegotiation (efficiency).Footnote 17 It does so, first, by attributing the right to participate in the renegotiation to two categories of Members: (a) initial negotiation rights (INR) holders and (b) countries with a principal supplying interest (PSI). Second, it attributes a right to be consulted (without a legal right of participation) to any Member with a substantial (supplying) interest (SI). Furthermore, and crucially, any of the Members identified above have the right to withdraw substantially equivalent concessions in the case where they are not happy with the new tariff concession (Art XXVIII:3). It is thus clear that only the Members that fall under any of the three above-mentioned categories will be able to protect their export interests in the context of a request for renegotiation by the applicant member. In other words, falling under any one of those categories provides members a stairway to heaven.
China's claim that the EU had violated Article XXVIII:1 by failing to negotiate or consult with it was premised on China demonstrating that it was either a PSI or SI country.
The Interpretative Note of Article XXVIII:1 GATT clarifies that the term PSI country typically covers a Member that ‘has had, over a reasonable period of time prior to the negotiations, a larger share in the market of the applicant contracting party than a contracting party with which the concession was initially negotiated or would, in the judgement of the contracting parties, have had such a share in the absence of discriminatory quantitative restrictions maintained by the applicant contracting party’ (para. 4).
The Interpretative Note of Article XXVIII:1 GATT also clarifies that the expression ‘substantial interest’ should ‘cover only those contracting parties which have, or in the absence of discriminatory quantitative restrictions affecting their exports could reasonably be expected to have, a significant share in the market of the contracting party seeking to modify or withdraw the concession’ (para. 7).
In the context of Article XXVIII:1, a 10% import share benchmark has been applied for the purpose of determining which Members hold a substantial supplying interest.Footnote 18
While up to 2008 China's share of imports into the EU of the product at issue was zero, from 2009 China's share increased exponentially at least for two of the products: China's share of imports of tariff line 1602 39 29 grew from 29% in 2009 to 52.8% in 2011 and imports of tariff line 1602 39 80 from 8.7% in 2009 to 61.1% in 2011 (when the EU concluded its renegotiation with other Members).Footnote 19
China's claim that it should have participated in the renegotiations pursuant to Article XXVIII:1 was based on two main arguments. First, China argued that the EU should have determined which Members held a principal or substantial supplying interest taking into account the fact that imports of the relevant products from China were subject in the relevant years preceding the two rounds of renegotiation (2003–2005 and 2006–2008) to SPS measures imposed by the EU. In other words, these SPS measures on poultry imports from China were ‘discriminatory quantitative restrictions’ within the meaning of paragraphs 4 and 7 of the Interpretative Note to Article XXVIII:1 (excerpted above).
The Panel rejected China's first argument, concluding that the SPS measures imposed by the EU on poultry imports from China did not qualify as ‘discriminatory quantitative restrictions’ for purposes of determining which Members held a principal or substantial supplying interest. While it accepted that the term ‘discriminatory’ ‘is somewhat elastic and may be interpreted narrowly or broadly, depending on the context’, the Panel nevertheless concluded that the ‘ordinary meaning of the term “discriminatory quantitative restrictions” would still only cover … quantitative restrictions that draw distinctions between imports from different countries that are similarly situated’.Footnote 20 Applying this concept of discrimination to the SPS measures, the Panel considered that ‘restrictions applied to imports based on sanitary grounds are “discriminatory” … only if imports from different countries that are similarly situated in terms of the sanitary situation or sanitary risks are not similarly restricted’.Footnote 21 The Panel buttressed its narrower interpretation of the term ‘discriminatory’ pointing to the complexity of the counterfactual analysis required to determine which Members hold a principal or substantial supplying interest if such determination had to be based on a broader interpretation of the term ‘discriminatory’ and thus on a wider universe of quantitative restrictions.Footnote 22 In this regard, the Panel referred to one of the objectives of Article XXVIII:1 expressly noted in paragraph 4 of the Interpretative Note to Art XXVIII:1, that is to ensure that the negotiations and agreement under Article XXVIII are not ‘unduly difficult’ and that ‘complications in the application of this Article’ are avoided.Footnote 23
The second argument advanced by China to support its claim that it should have participated in the renegotiations pursuant to Article XXVIII:1 revolves around whether the EU was obliged to re-determine which Members held a supplying interest to reflect changes in import shares that took place following the initiation of the negotiation. China's argument was based on the fact that, while the EU notified its intention to modify its concession in 2009, the negotiations were not concluded until 2012. Crucially, as noted above, China's share of imports into the EU of the products at issue increased exponentially at least for two of the products in the period between 2009 and 2011. The Panel rejected China's argument mainly based on the ‘absence of any guidance on’ whether an applicant member is required to reappraise which Members held a supplying interest to reflect any changes in import shares following the initiation of the negotiations.Footnote 24 The Panel concluded that in light of this silence, ‘and the need to strike a delicate balance between the different objectives of Art XXVIII …, we cannot, as treaty interpreters, formulate a general rule on this matter’.Footnote 25 Furthermore, while it recognized that there is a six-month time limit in the case of negotiations under Article XXVIII:1, the Panel stated that ‘in the case of reserved negotiations pursuant to Art XXVIII:5, there are no time-limits specified regarding when such negotiations are to be concluded’.Footnote 26
Accordingly, two factors seem to have played a key part in China's failure to convince the Panel that the EU should have considered China as holding a principal or substantial supplying interest: first, the prevalence of the need to avoid complexity and thus guarantee efficiency in the renegotiations (over the need to guarantee fairness) and, second, the unwillingness to fashion an unwritten rule in the absence of any textual guidance. The former was key to rejection of China's first argument that ‘discriminatory quantitative restrictions’ should have included the SPS measures imposed by the EU, while the latter was key to rejection of China's second argument that the EU should have reassessed China's share of imports in light of the length of the negotiations.
We have a few concerns with regard to the Panel's reasoning. First, we question the wisdom of the Panel's focus on the ‘object and purpose’ of the specific provision at issue rather than on the several objects and purposes of the overall treaty under review (including those stemming from the specific provisions at issue) as provided by the well-known customary rule on treaty interpretation. Over emphasis on the object and purpose of a single provision may lead to losing sight of the bigger picture.
Second, the reasoning behind the two findings may also appear contradictory, as relying on the underlying objectives of Article XXVIII would logically allow the interpreter, at least to some extent, to fill in some of the gaps in the specific disciplines provided for in Article XXVIII and related documents. For example, if the prevailing aim is a speedy and not-too-complex renegotiation of existing tariff binding, then one can envisage the situation where a time limit for concluding such renegotiation should be read into the rules. And the same may be said for the reverse (a purely textual reading of the term ‘discriminatory’) (for a critique of the textualist approach see Horn and Weiler, Reference Horn, Weiler, Horn and Mavroidis2005).
Third, aside from the possible contradiction, it seems apparent that the interpretative exercise could have, in principle, reached a different outcome, one which would have ultimately recognized China as having the status of a country with a principal, or substantial supplying interest, and thus allowed China to effectively take part in the renegotiation or consultation with the EU.
3.2 Re-allocating TRQs under Article XIII:4: an empty obligation?
An interesting aspect of the Panel's decision on the allocation of TRQs under Article XIII GATT focuses on the question whether Article XIII:4 GATT imposes an obligation to reallocate TRQ shares upon request from a Member with a substantial interest in supplying the product subject to country-specific quotas. This appears to be an important safeguard in the ability of Members with growing export capabilities to challenge existing country-specific quotas. While Article XIII:2(d) does allow for some flexibility in this regard as it requires that ‘any special factors’ be taken into due account while allocating quotas to countries with substantial supplying interest,Footnote 27 it is principally geared to respect historical shares rather than the non-discriminatory administration of quotas.Footnote 28
While the Panel found that Article XIII:4 can be interpreted as establishing a legal obligation on the importing Member to reallocate quota (including TRQ) shares among supplying countries (to reflect changes in the import shares held by different countries), the Panel concluded that ‘there is no obligation to reallocate the TRQ shares within any particular time frame, and at least not in the years immediately following the initial TRQ allocation’.Footnote 29 Accordingly, in the Panel's view, the EU did not violate Article XIII:4 when it refused to reallocate the TRQ allocations (arising from the second modification package) in May 2014.
The reasoning is, to say the least, problematic. The relevant provision states in part as follows: ‘the contracting party shall, upon the request of any other contracting party having a substantial interest in supplying that product … consult promptly with the other contracting party … regarding the need for an adjustment of the proportion determined’. The Panel could have decided, as argued by the EU, that Article XIII:4 only provides for a consultation obligation imposed on the contracting party imposing the country-specific quotas, but it does not include an obligation to reallocate the quota shares based on changes in import shares. However, citing dicta from previous Appellate Body decisions (dealing with the EU banana import regime), the Panel noted that, while ‘the Member imposing and allocating the TRQ must have a degree of discretion as to whether or not it should reallocate the TRQ shares following a request for consultation under Article XIII:4’, it did not consider that ‘this discretion is unfettered, such that a Member maintaining a TRQ is free to ignore significant changes in imports shares held by different countries following the opening of a TRQ’.Footnote 30
This appears to be a bold decision, certainly based on the earlier (over-) reliance on the text of the GATT (i.e., the text does not appear to set an obligation to reallocate). However, what is really surprising is that the Panel does not offer any rationale for why such discretion is not otherwise unfettered. There is no reference, for example, to the ‘object and purpose’ of the specific provision at issue (Article XIII:4), something the Panel has used previously in the context of interpreting Article XXVIII.
However, this is as far as the Panel's boldness reached. Having proclaimed the existence of limits to the Member's discretion to refuse to reallocate the TRQ shares, the Panel was not able to identify ‘any time frame as to when or how often such reallocation would have to take place, or based on the occurrence of which events’.Footnote 31 The Panel once again relied on the silence of the relevant provision (‘no specific guidance in the text’),Footnote 32 ‘China's inability to clearly specify the time frame, frequency and the basis upon which the redetermination of the TRQ allocation should be made’,Footnote 33 and the overall ‘prevalence and centrality of historical market shares in TRQ share allocations’.Footnote 34
We believe this shows not only the weakness of (and contradictions within) the Panel's reasoning underlying its interpretation of Article XIII:4, but also the existence of an alternative reading of this provision, which would provide the basis for a substantive right of reallocation of relevant quotas (according to parameters that would need to be established) to reflect changes in the import shares held by other countries.
4. Economic analysis
Economic analysis can provide essential guidance in determining a principal supplying interest for the renegotiation of schedules under Article XXVIII and the allocation of TRQs under Article XIII. Ultimately, we want to understand the general tradeoffs in using a more inclusive definition of principal supplying interest (or equivalently, substantial interest), and how these tradeoffs apply in the current case. Such an analysis requires us to answer broader questions on the economics of the object and purpose of the WTO, and how TRQs and renegotiations serve that object and purpose.
This section first provides background on the economic effects of the EU's modifications. We conclude that the EU's policies effectively impose discriminatory restrictions on nations other than Brazil and Thailand, while causing trade diversion from potentially more efficient suppliers. Though not legally relevant, the EU's primary policy objective we infer to be the creation of economic surplus or rents for Brazil and Thailand, as part of a mutually agreed solution to an earlier pairing of poultry meat disputes that Brazil and Thailand won against the EU. We further discuss the dynamic inefficiencies created as developing nations increase their capacity for poultry exports while facing rigid TRQs, so the economic analysis further supports the right to TRQ reallocation from the legal analysis.
As for the general question of how inclusive the principal supplier rule should be, we argue that a relatively inclusive and flexible definition of principal supplying interest is worthwhile in the current context of modifying tariffs to TRQs. We argue that modifications from tariffs to TRQs, though explicitly permitted in the interpretive note of Article XXVIII:1, do not fit well with the broader purpose of the WTO. Such modifications can be economically equivalent to new discriminatory tariffs. Based on the tradeoffs involved, substantial interest in Article XXVIII renegotiations of schedules should be interpreted more inclusively (as permitted by paragraph 7 of the interpretive note of Article XXVIII:1 cited previously), in part because allowing more nations to withdraw equivalent concessions discourages the unwinding of tariff liberalization, and thus better serves the overall object and purpose of the WTO agreements. We conclude by suggesting that China's claims as a principal supplier for renegotiations could have been better argued using better economic methods.
4.1 The economic effects of the EU's modifications
We first review the static and dynamic economic effects of the TRQ in isolation. Schropp and Palmeter (Reference Schropp and Palmeter2010), in an earlier commentary on EC–Bananas III, provide an excellent discussion of the economics of TRQs. For analysing the poultry market, the standard perfectly competitive model of TRQs is a reasonable approximation. The first observation about TRQs is that there are three types of competitive equilibria depending on import demand and the rates set by the EU. Sufficiently low demand would imply that the TRQ is no different from a tariff at the lower in-quota rate. For a large intermediate range of demands or prohibitively high out-quota rates, the TRQ effectively functions as a quantitative restriction. The last possibility is a two-tiered system such that favoured importers trade at the in-quota rate and the rest of the world trades at the out-quota rate.
The ultimate effect of the EU's modifications was to reduce tariffs faced by the principal suppliers Brazil and Thailand, while substantially raising tariffs on the rest of the world, including China. Though Article XIII may be titled ‘Non-discriminatory administration of quantitative restrictions’, the TRQs are still discriminatory in an economic sense (Schropp and Palmeter, Reference Schropp and Palmeter2010, make similar points). Because the policy is effectively discriminatory, it can create trade diversion, which occurs when policy displaces a more efficient supplier of any export unit with a more inefficient supplier. Even if Brazil and Thailand are more efficient suppliers for most units, in a world where each country has an increasing supply curve, even small countries could be capable of supplying some goods to the world market absent tariffs. Their exclusion is a source of static inefficiency. Economic theory then suggests some sense in China's claim under Article I that the TRQs violate non-discrimination, though legally the TRQs are clearly governed by Article XIII.
There are some subtle differences between trade diversion in the case of TRQs and the more standard textbook case of trade diversion due to discriminatory tariffs (e.g. Krugman, Melitz, and Obstfeld, Reference Krugman, Obtsfeld and Melitz2014). Typically, the welfare implications of a discriminatory tariff reduction reflect a tradeoff between trade creation (from reducing a tariff) and the aforementioned trade diversion. For a particular TRQ though, there clearly is no trade creation, because the trade volume is fixed. So there is only trade diversion. Thus, there is no welfare benefit for the EU by allowing less efficient suppliers at in-quota rates.
Why then would the EU undertake such a policy that in isolation leads to a welfare loss? The answer must lie in the EU benefiting from distributing producer surplus or quota rents to Brazil and Thailand. The policy was undertaken as part of a mutually agreed solution to Brazil and Thailand from past poultry disputes lost by the EU.Footnote 35 As an aside, notice then that the current case then exemplifies concerns often raised about how mutually agreed solutions to WTO disputes can create new discrimination that begets further disputes, rather than removing the offending policies (Evenett and Jara, Reference Evenett and Jara2014).
In addition to the static considerations mentioned, dynamic efficiency is economically important in this case. China's productive capacity in poultry naturally grew relative to the rest of the world after its WTO accession. Absent reallocations, a TRQ with a prohibitive out-quota rate would lead to increasing inefficiency as China continues to grow. Though tariffs and quotas may be equivalent for a given trade volume in static perfectly competitive markets, for a growing exporter the quotas are an inferior instrument, because they lock in a particular trade volume. The failure of quotas to adjust to growth, the political problems of allocating quota rents, and the relative ease of negotiating reciprocal tariffs are among the many reasons for the GATT (and later the WTO) to encourage the process of tariffication: converting quantitative restrictions to tariffs.
Aside from distributing surplus to Brazil and Thailand, the EU's policy could reflect a longer-run desire to protect local producers from Chinese poultry imports. Rather than modifying tariffs to TRQs for more permanent protection, economists would instead encourage temporary trade protection, particularly non-discriminatory safeguards, to aid the transition. Nonetheless, the remainder of this section will evaluate whether a more permanent modification of concessions from tariffs to TRQs still somehow fits into the broader object and purpose of the WTO.
4.2 Economic view of the object and purpose of tariff-rate quotas
Before assessing the principal supplier rule and TRQs, we need to first specify a political economic view of the object and purpose of the WTO and its agreements. Modern political economic theory of the WTO (Bagwell and Staiger, Reference Bagwell and Staiger1999), views the organization as primarily addressing international externalities that arise through the policy choices of nations, as well as providing an avenue for gradual domestic commitment from political pressure.Footnote 36 This body of theory acknowledges that political economy will influence government preferences, and an ultimate goal is to achieve policies that are Pareto efficient given those preferences. The theory argues the WTO is designed to pick not just any Pareto efficient outcome, but rather one that is power-neutral and rules-based rather than power-based. Notice that such an outcome need not be free trade. The principles of reciprocity and non-discrimination facilitate gradual liberalization towards such a politically optimal outcome.
Theory tends not to explicitly address TRQs, however, so how well do they fit into such a framework? The TRQ was considered as a gradual step between pure quantitative restrictions and pure tariffication by Uruguay Round negotiators (Downes, Reference Downes2017), so a starting point is to consider how TRQs fit into economic theories of gradualism. A typical theory of gradualism allows some frictions in the transition under trade liberalization, such that once there is more adjustment in the economy, further trade liberalization becomes possible. Footnote 37
There are reasons to doubt this sanguine view of TRQs as an effective form of gradualism. As noted by Schropp and Parmenter (Reference Schropp and Palmeter2010) in the context of EU–Bananas III, the historical allocation of quotas tends to lock in current export suppliers for political reasons rather than encourage a dynamic progression towards global efficiency. Indeed, the record of TRQs encouraging a progression to tariffs in agriculture following the Uruguay Round is rather poor, and instead TRQs have created another avenue for rent-seeking (Downes, Reference Downes2017).
4.3 Economic view of the object and purpose of renegotiations
Renegotiations allow for modifications of concessions, as long as trading partners can withdraw equivalent concessions. Multiple economic models view renegotiations of schedules as serving the broader efficiency-enhancing mission of the WTO, though these models apply less well to modifications of tariffs to TRQs.
In a static setting, renegotiations can still guide nations closer to a more power-neutral outcome, even if the renegotiation in isolation were to decrease efficiency. A country in a weaker bargaining position can then demand renegotiations, through which the reciprocal withdrawal of concessions then leads to a better balance of welfare between governments. Renegotiations effectively collapse the Pareto efficiency frontier into a set of points that is overall closer to the power-neutral political optimum (Bagwell and Staiger, Reference Bagwell and Staiger1999, in particular Figure 5B). We note though that a key feature of this theoretical argument is that the withdrawal of concessions is reciprocal.
We can also consider renegotiations using the economics of liability rules. Tariff concessions in the WTO are like liability rules, which can be broken if compensation is provided to the affected party, as opposed to property rules, for which punishment is designed to be prohibitive. In the WTO, nations can unilaterally break commitments, and members can then withdraw equivalent concessions through the renegotiation process. As trade conditions evolve in a dynamic setting, the WTO permits a form of efficient breach: nations will only violate agreements if they value their modification of concessions more than the equivalent concessions withdrawn by trading partners. Such a breach should then improve efficiency, though it has been relatively uncommon in the WTO (Pelc, Reference Pelc2010), so renegotiation obligations have primarily served as a deterrent to modifying concessions. The fact that nations do not make direct transfers to one another in the WTO complicates the efficient breach argument, but by and large it still holds (Maggi and Staiger, Reference Maggi and Staiger2015) As with the previous economic theory, the key to the efficient breach argument is that the withdrawal of concessions is reciprocal.
The above theories though do not apply well to the modification of tariffs to TRQs, because such modifications do not involve reciprocal withdrawals of concessions. In the current case, Brazil and Thailand were granted tariff-rate quotas mostly at the previous bound rate and some at lower rates, which were all gratefully accepted without further modifications. In contrast, all countries excluded from renegotiations and TRQ allocations were not entitled to any withdrawal of concessions. There is then no disincentive for nations to modify such tariffs to TRQs, so we cannot say the EU modifications fit into the model of efficient breach. Nor can we claim that this modification is guiding nations to a more power-neutral outcome given that the EU suffers no consequences from the modification.
Since the tariff-to-TRQ modification does not contribute to the WTO's object and purpose, we can conclude that the modification effectively creates a new discriminatory barrier on all countries not deemed to be principal suppliers. The analysis then motivates a more permissive view on how principal suppliers should be determined, in order to facilitate more efficient and power-neutral outcomes.
4.4 Economic views on the object and purpose of the principal supplier restrictions
The economic analysis thus far considers only the benefits of identifying more principal suppliers. But there are costs and tradeoffs involved, as evidenced by the design of Article XXVIII to limit the number of countries involved in renegotiations. What then does economics imply about why principal suppliers should be limited? Should these limitations be more or less stringent in the context of tariff-to-TRQ modification?
Recent work on the economics of the principal supplier rules argues that MFN free-riding is central to the need to limit the numbers of principal suppliers. Though MFN free-riding does not exist when all tariff reductions follow principles of reciprocity and non-discrimination (Bagwell and Staiger, Reference Bagwell and Staiger1999), researchers find that MFN free-riding exists in practice. According to Ludema and Mayda (Reference Ludema and Mayda2009), without the principal supplier rule, importers would have too little bargaining power facing many exporters wanting to free ride off the market access granted by MFN tariff reductions. The principal supplier rule ultimately allows the importer's participation constraint in negotiations to be satisfied by limiting the bargaining power of exporters. The empirical results of Ludema and Mayda (Reference Ludema and Mayda2013) support the relevance of the principal supplier rule in explaining Uruguay Round tariff negotiations.
The theory speaks less well to the value of the principal supplier rule in the context of renegotiations. According to textbook economic theory, the involvement of more countries can complicate negotiations, since any one country could attempt to ‘hold up’ negotiations. But absent the MFN free-rider problem, we conclude that the argument for limiting principal suppliers is stronger for negotiations than renegotiations.
Relatedly, we discuss the concern that China may be free-riding in the renegotiation, as it did not make initial concessions itself to gain market access to the EU. First, notice that Article XXVIII explicitly includes the language of principal suppliers to allow the possibility that not all principal suppliers may have made concessions themselves. Second, notice that China negotiated an accession protocol that allowed it to achieve the market access granted in poultry meat among many markets, and thus it made concessions. China should then have reasonably expected the opportunity to become a principal supplier in poultry meat.
4.5 Using economics to assess principal supplying interests
To conclude, the preceding economic analysis suggests that renegotiations involving TRQs would benefit from a much more inclusive view of which suppliers are included. In the actual ruling, the EU used a benchmark threshold of 10% import share based on a particular reference period, and the Panel ruled this benchmark to be acceptable in allocating TRQs once special factors were considered. For the purpose of renegotiations, the Panel ruled out any obligation for the EU to consider China's export potential, which had been limited by the SPS measures in place during the reference period. All arguments China brought in defence of its principal supplying interest for the purposes of renegotiations were disregarded. Our previous legal and economic analysis suggests the Panel should accept a broad range of evidence in support of identifying a principal supplying interest for renegotiations of schedules.
In its claims, China offered as evidence its domestic production of poultry and exports to other markets during the reference periods for renegotiations, as well as its eventual exports to the EU once the SPS measures were relaxed. The EU dismissed all these additional trade data as irrelevant in its determination of principal suppliers for renegotiations under Article XXVIII, and the Panel did not find the EU to be in violation of its obligations. But modern economic analysis can indeed utilize such information to provide counterfactuals for China's exports absent SPS measures during the relevant reference periods.Footnote 38 We have not conducted this analysis ourselves, but such a method could have potentially been used to justify China's inclusion in renegotiations of schedules under Article XXVIII in the two tariff lines in which China eventually achieved 50% import share.
By ruling that China's import increase was an exceptional special factor that entitled China to a TRQ allocation, the Panel came to a similar conclusion as a counterfactual analysis might have in terms of assessing an EU obligation to China. But using a counterfactual analysis to determine China's principal supplying interest in the renegotiation of schedules would have been a more robust economic approach. The Panel's failure to recognize any EU obligation to China in the renegotiation of schedules leaves room for concern about whether future modifications of tariffs or TRQs will occur without recognizing obligations to WTO members who are substantially affected by the modifications.
