The EC – Tariff Preferences Case

The EC – Tariff Preferences case dealt with conditionalities in the European Union’s (EU) ^{Footnote 23} GSP. It was the first case in the WTO in which the Enabling Clause was adjudicated. ^{Footnote 24} It interpreted the meaning of non-discrimination and proved that the principle of non-discrimination, being the cornerstone of the multilateral trading system, must be complied with even in the exceptions, which in this case was the Enabling Clause. This means that there is really no exception to the principle of non-discrimination. In this case, India challenged Council Regulation (EC) 2501/2001 Applying a Scheme of Generalised Tariff Preferences (Regulation 2501/2001), which laid down the following five schemes of preferences:

1. 1. general arrangements;

2. 2. special incentive arrangements for the protection of labour rights;

3. 3. special incentive arrangements for the protection of the environment;

4. 4. special arrangements for least developed countries; and

5. 5. special arrangements to combat drug production and trafficking. ^{Footnote 25}

The preferences under the general arrangements were granted to all developing countries, whereas the fulfilment of certain conditions was required to benefit from additional preferences under the special arrangements. The special incentive arrangements for the protection of labour rights and the environment were so stringent that most developing countries did not even apply for them, and only two succeeded in fulfilling the conditions. ^{Footnote 26} The advantages under the last category — that is, the drug arrangements — were granted to eleven countries whose products benefited from zero tariffs, ^{Footnote 27} whereas the products of other developing countries benefited from reduced tariffs or had to pay the entire tariff. ^{Footnote 28} The EU felt that the eleven beneficiaries needed these preferences to stimulate their economic growth and improve their commercial possibilities so that the citizens of these countries would abandon the manufacture of illicit drugs and take up the manufacture of licit products instead. ^{Footnote 29}

In 2001, the EU added Pakistan as the beneficiary of its additional preferences under the drug arrangements. The European Commission acknowledged that the EU granted additional preferential market access to Pakistani textiles and clothing to reward Pakistan for its position against the Taliban and also to gain access to the Pakistani market. ^{Footnote 30} The grant of additional preferences to Pakistani textiles and clothing led to a distortion of the conditions of competition between India and Pakistan. Thus, it had a negative impact on India’s exports since India and Pakistan were competitors in the export of textiles and clothing to the EU. ^{Footnote 31} In addition, developing countries that lost market access such as India paid for the increased market access for other developing countries such as Pakistan. ^{Footnote 32} Therefore, greater trade preferences can benefit a country at the cost of its competitor. In other words, increased market access can eradicate the problem of drugs in a country, but reduced market access can increase the same problem in another country. Obviously, the aim of the Enabling Clause is not to transfer the problems of one country to another, but this is exactly what the drug arrangements led to. ^{Footnote 33}

India challenged the drug arrangements, arguing that they violated the MFN treatment principle and were not justified by the Enabling Clause because footnote 3 ^{Footnote 34} prevented the GSP donors from granting non-identical preferences to their beneficiaries. Moreover, paragraphs 2(a) ^{Footnote 35} and 3(c) ^{Footnote 36} referred to all developing countries. Since paragraph 2(a) did not allow donors to select their beneficiaries, preferences could not vary due to the needs of developing countries. So India’s argument was that developing countries had not given up on the MFN treatment principle, which applied even when preferences were granted and, thus, prevented donors from distinguishing between beneficiaries. In other words, India was arguing that the principle of non-discrimination applies even in the exceptions. Normally, the word “discrimination” carries a negative connotation and means unjustified distinction. However, India was arguing that any distinction, even if justified, amounted to discrimination and that GSP schemes must benefit all developing countries, without differentiation, to be protected by the Enabling Clause.

The EU argued that the Enabling Clause excluded the application of the MFN treatment principle. It stated that the drug arrangements were covered by paragraph 2(a) of the Enabling Clause instead of by the MFN treatment principle. It also argued that “non discriminatory” in footnote 3 of the Enabling Clause did not require identical preferences to be granted and allowed distinctions based on objective criteria such as developing countries’ needs. Moreover, paragraph 2(a) did not mention “all” developing countries, and this was confirmed by paragraph 3(c), which allowed preferences suited to the needs of developing countries.

Despite the important stakes India had in this case, Pascal Lamy, who was the European trade commissioner at that time, remarked that India did not need preferences, suggesting that India would not benefit from the case. ^{Footnote 37} However, India chose to bring the case not only due to economic reasons but also because it raised a systemic question regarding the conditionalities in the GSP. ^{Footnote 38} Conditionality can be defined as granting benefits subject to the beneficiary meeting certain conditions. However, it carries a stronger meaning, as a mechanism to bring about policy reform in the beneficiaries or to impose policies that the beneficiary would not choose voluntarily. ^{Footnote 39} In the World Bank, it implies a provision of financial support to beneficiaries in return for the implementation of structural changes. ^{Footnote 40} In other words, conditionality is related to power since the balance of power between the donor and the beneficiaries is important. ^{Footnote 41}

The EU provides additional trade preferences to countries that comply with certain conditions. The EU’s rationale behind this is that “economic benefits are privileges to be granted to developing countries that comply with democratic principles and human rights, and to be withdrawn from those that do not.” ^{Footnote 42} However, the EU does not include human rights and democracy conditionalities in its relations with developed countries. ^{Footnote 43} Of course, these relations exclude the GSP. But the EU’s difference in treatment of developing and developed countries shows that conditionalities are related to power.

India’s complaint raised two related questions. The first question related to the type of differentiation authorized — preferences granted to beneficiaries must be identical or is it possible to distinguish between them on the basis of different criteria? ^{Footnote 44} The second question related to the status of footnote 3 in the Enabling Clause — was it binding and how to interpret it? ^{Footnote 45} The panel found that the drug arrangements violated the MFN treatment principle and were not justified under paragraph 2(a) of the Enabling Clause. It also found that the term “non discriminatory” in footnote 3 of the Enabling Clause required identical preferences to be granted, without differentiation, to all developing countries; however, it allowed a priori limitations. ^{Footnote 46} Additionally, it found that the term “developing countries” in paragraph 2(a) of the Enabling Clause referred to all developing countries, but allowed a priori limitations. ^{Footnote 47} The panel referred to UNCTAD Resolution 21(II) to support its finding that the term “non discriminatory” meant a complete absence of distinction including on the basis of objective criteria. In fact, it found no objective criteria to distinguish between different development needs such as drug trafficking, poverty or poor education, and so on. ^{Footnote 48}

The EU appealed and argued that the term “non discriminatory” obliged members to grant objective preferences because paragraph 3(c) required a response to the needs of developing countries. Moreover, paragraph 2(a) did not refer to all developing countries since paragraph 3(c) allowed objective distinctions. The Appellate Body found that the term “non discriminatory” in footnote 3 of the Enabling Clause constituted an obligation and that GSP schemes would have to be non-discriminatory to be justified under paragraph 2(a) of the Enabling Clause. ^{Footnote 49} However, the term “non discriminatory” did not require identical preferences to be granted. ^{Footnote 50} Moreover, the term “developing countries” in paragraph 2(a) of the Enabling Clause did not refer to all developing countries. ^{Footnote 51} The Appellate Body also stated that different developing countries could have different development needs susceptible of changing because development did not happen in a uniform manner in all developing countries. ^{Footnote 52} It continued in the same vein by stating that developing countries could have different needs in accordance with their levels of development and particular circumstances. ^{Footnote 53} The Appellate Body referred to the preamble to the Agreement Establishing the World Trade Organization (WTO Agreement), which refers to the needs of members at different levels of economic development, to support its finding. ^{Footnote 54}

The Appellate Body used paragraph 3(c) of the Enabling Clause as context to interpret footnote 3 ^{Footnote 55} and stated that “we read paragraph 3(c) as authorizing preference-granting countries to ‘respond positively’ to ‘needs’ that are not necessarily common or shared by all developing countries. Responding to the ‘needs of developing countries’ may thus entail treating different developing country beneficiaries differently.” ^{Footnote 56} It continued:

Applying this criterion to the drug arrangements, the Appellate Body found that they were not available to all the beneficiaries of the GSP that were facing the problem of drug production and trafficking. The drug arrangements were limited to twelve developing countries, and indicated neither the manner of selecting the beneficiaries nor the elements for determining the effect of the drug problem. There was no way to add other countries as beneficiaries of the arrangements. Furthermore, there were no criteria to distinguish beneficiaries of drug arrangements from other beneficiaries of the EU GSP. The drug arrangements did not lay down the criteria to remove beneficiaries. This meant that a beneficiary could continue to benefit from the drug arrangements irrespective of whether or not the arrangements resolved the drug problem in the country. ^{Footnote 58} This would certainly not be very encouraging for the development of developing countries facing drug problems that could not benefit from the drug arrangements.

Moreover, Regulation 2501/2001 did not state the method of evaluating the drug arrangements’ response to the drug problems. This shows that the drug arrangements were opaque and were also applied in a discriminatory fashion. Additionally, the EU had asked for a waiver from the MFN treatment principle because the drug arrangements benefited the products of twelve countries only. ^{Footnote 59} Therefore, the Appellate Body did not agree with the EU that all developing countries were potential beneficiaries of the drug arrangements. Consequently, the EU was unable to prove that the drug arrangements satisfied the requirement of “non discriminatory” in footnote 3 and that they were justified under paragraph 2(a) of the Enabling Clause. ^{Footnote 60}

Evaluation of the EC – Tariff Preferences Case

This case not only raised a number of difficult questions but also left quite a few of them unanswered, thus leading to uncertainty regarding GSP schemes. The main reason for this lack of clarity is that the Enabling Clause is ambiguous, and this was acknowledged by the panel and the Appellate Body. In fact, it does not provide definitions for the terms “non discriminatory” and “developing countries.” The Enabling Clause refers neither to all developing countries nor to particular developing countries. ^{Footnote 61} Like any decision, this one also has its pros and cons. Consequently, there have been varied reactions to the case since it was important for GSP donors as well as for beneficiaries. Some developing countries felt that the Appellate Body’s finding would fragment their unity, while others felt it reflected the diversity among developing countries. ^{Footnote 62} The decision in this case can be perceived as disadvantageous for developing countries or as a compromise between different interests. ^{Footnote 63}

Brazil felt that the Appellate Body legitimized the use of the GSP as a tool of foreign policy. ^{Footnote 64} It can also be said that India won the case since the drug arrangements were struck down, but the interpretation of the law was in the EU’s favour. ^{Footnote 65} Both of the parties expressed their reactions to the decision. India expressed its dissatisfaction at the decision by stating that

Meanwhile, the EU stated in a press release that the decision was a victory for GSP donors, including the EU, which wanted to respond positively to the particular needs of the sub-groups of similarly situated developing countries. ^{Footnote 67}

It may be useful to examine the concept of discrimination since the entire case hinged on it. Non-discrimination in the WTO is not a linear concept. It varies from one legal provision to another. ^{Footnote 68} Moreover, conditionalities can lead to discrimination, but they can also be non-discriminatory. Thus, discrimination and conditionality are not synonymous. The condition can lead to discrimination when applied in a uniform manner to all beneficiaries. The requirement of fulfilling objective minimum conditions can be discriminatory since certain potential beneficiaries/developing countries that are not in a position to fulfil them may not benefit from the additional preferences. This means that the capacity of the potential beneficiaries to fulfil the conditions must be taken into account. ^{Footnote 69} Thus, non-discrimination includes treating unequals unequally. ^{Footnote 70} Lassa Oppenheim’s International Law: A Treatise explains discrimination as treating differently those who are in the same situation or treating in the same way those who are in different situations. ^{Footnote 71} Thus, countries situated differently can be treated differently. The Appellate Body also clarified this when it stated that similarly situated beneficiaries should be treated alike. This is also highlighted in Article XX of the GATT, ^{Footnote 72} which allows non-discriminatory conditions, Article 2.3 of the Agreement on the Application of Sanitary and Phytosanitary Measures, ^{Footnote 73} and paragraph 2(d) of the Enabling Clause. ^{Footnote 74}

The Appellate Body struck down the drug arrangements due to a lack of transparency, but it could be argued that the Enabling Clause does not require transparency in the administration of GSP schemes because a lack of transparency does not necessarily amount to discrimination even in the case of a closed list as long as the GSP donor has correctly evaluated the needs of the beneficiaries and the responses to those needs. ^{Footnote 75} However, this argument ignores the fact that closed lists cannot exist in the case of proper evaluation of needs, and the responses to those needs, that can only be made by practising transparency. The element of discrimination could have been removed by following this procedure. Since it was not followed, the drug arrangements were discriminatory because of a lack of transparency.

The Appellate Body’s decision allows more freedom to donors to distinguish between beneficiaries as long as these distinctions fulfil the requirements laid down in the Enabling Clause ^{,Footnote 76} as interpreted by the Appellate Body. The Appellate Body stated that these distinctions must be based on objective criteria such as those recognized in the WTO Agreement or in international instruments. ^{Footnote 77} As a consequence, the Appellate Body enhanced the possibility for donor countries to add conditions, including non-economic ones. This is obviously not favourable from the point of view of the potential beneficiaries that cannot always fulfil these conditions.

The Appellate Body simply assumed that the criteria it referred to were objective and did not define the term “objective.” It did not specify who bore the burden of proof to prove the objectivity of the criteria. ^{Footnote 78} In addition, it did not specify the method of application of the objective criteria. Should they be applied to the entire economy of the potential beneficiary or to its sectors or to the international standards such as the UN Human Development Index? ^{Footnote 79} This allows the GSP donor to select the method of application of the objective criteria. Moreover, the freedom to objectively distinguish may be misused to favour allies ^{Footnote 80} or to pressure developing countries in multilateral or bilateral negotiations. Donors may include these conditions in bilateral agreements with developing countries, ^{Footnote 81} and this would exclude recourse to WTO dispute settlement.

Moreover, one of the difficulties in implementing the criteria laid down by the Appellate Body will be to find out if potential beneficiaries are “similarly situated,” since the Appellate Body did not provide a definition of the term. In addition, there may be a number of needs/problems that need to be resolved, such as health, education, and so on, and it may be discriminatory if the EU only grants preferences to resolve drug problems as opposed to other needs. ^{Footnote 82} The unfettered discretion to choose the needs to be addressed may amount to de facto discrimination among beneficiaries, thus highlighting the weakness of the criteria laid down by the Appellate Body. The Appellate Body did not require a GSP donor to respond to all possible development, financial, and trade needs of the potential beneficiaries. ^{Footnote 83} And this fact leads to the following questions:

• • how is one to define a development, financial, or trade need;

• • do GSP donors have absolute discretion to choose the problems that they resolve;

• • how can the problem be quantified in order to be resolved;

• • is exclusion of sensitive products from the benefits of GSP schemes not discriminatory; and

• • do these conditions not amount to a requirement of reciprocity? ^{Footnote 84}

Additionally, the economic interests of beneficiaries evolve. A broad GSP scheme covering many products (espoused by the panel ^{Footnote 85} ) allows the beneficiaries to shift to exporting newer products. A narrow GSP scheme tailored to similarly situated beneficiaries does not allow the beneficiaries to shift to exporting newer products because this shift is governed by the GSP donor who might have an interest in maintaining the beneficiaries in the existing position. Besides, it is possible that the beneficiaries that are unable to make the shift are actually inefficient producers of those products. ^{Footnote 86}

Another consequence of the Appellate Body’s ruling is that GSP donors will determine whether or not a country is a developing country, a determination that is normally made by the developing countries themselves. Potential beneficiaries will have to prove that they fulfil the conditions or that the conditions are unlawful. However, the Appellate Body has not clarified if the potential beneficiary has to prove that it fulfils the conditions or if the GSP donor has to prove that the potential beneficiary does not fulfil the conditions. ^{Footnote 87} These issues can be avoided in a GSP scheme without conditions.

Furthermore, there is a contradiction between the decision of the Appellate Body that makes “non discriminatory” a requirement and the Doha Decision on Implementation-Related Issues and Concerns that states that preferences “should” be non-discriminatory. ^{Footnote 88} Given the de facto rule of precedent that operates in WTO dispute settlement, it is possible that a future panel or Appellate Body would consider the decision in the present case. But the Doha Decision on Implementation-Related Issues and Concerns may also be relevant under Articles 31.3(a) and 31.3(b) of the 1969 Vienna Convention on the Law of Treaties (VCLT), which refer to subsequent agreement or practice of the parties regarding the interpretation of the treaty. ^{Footnote 89}

Paragraph 3(a) of the Enabling Clause states that “[a]ny differential and more favourable treatment provided under this clause: (a) shall be designed to facilitate and promote the trade of developing countries.” If the GSP donors are allowed to distinguish between potential beneficiaries on shaky objective grounds, would they not be contradicting this paragraph? So is the Appellate Body’s interpretation in conformity with this paragraph? ^{Footnote 90} The Appellate Body stated that differentiation was allowed to respond to the needs of developing countries. The drug arrangements were not based on any criteria and were thus discriminatory. India’s argument that no distinction should be made between beneficiaries based on their needs would lead to discrimination because granting identical preferences to developing countries that have different needs would not allow for a response to these different needs, thus leading to discrimination. Nevertheless, responding differently to different needs requires more complex laws. ^{Footnote 91} Therefore, the Appellate Body’s approach is harder to implement compared with that of the panel.

Footnote 3 and paragraph 3(c) have always been present in the Enabling Clause, but their status has been ambiguous. By interpreting them as requirements, the Appellate Body has placed limits on the freedom of GSP donors to impose arbitrary conditions. These limits are necessary to avoid arbitrariness, ^{Footnote 92} as was manifest in the EC – Tariff Preferences case. Theoretically, the Appellate Body’s interpretation is more favourable to potential beneficiaries since it would allow the donors to respond to needs requiring additional preferences. In reality, it will depend on the beneficiaries’ capacity to fulfil the conditions.

This case will have an impact on negative or positive conditionalities in GSP schemes that are difficult to fulfil and that distinguish between similarly situated beneficiaries. The Appellate Body stated that preferences should bring about a positive response to the needs of the beneficiary. ^{Footnote 93} How would negative conditions bring about positive responses unless withdrawal of preferences is seen as contributing positively to the alleviation of a problem or need? So the Appellate Body’s interpretation might push donors to change from negative to positive conditionalities and lead to a decrease in reasons that qualify as needs, justifying a distinction between the beneficiaries. ^{Footnote 94}

Certain scholars argue that the GSP will lead to de facto discrimination among beneficiaries since it is supposed to promote industrialization through the export of manufactured goods. Other factors leading to discrimination include the fact that some beneficiaries are unfit to be recipients politically and that sensitive domestic sectors of donors need to be protected. ^{Footnote 95} By way of example, the US GSP uses geopolitical considerations and the sensitivity of products to determine beneficiary status. ^{Footnote 96} Moreover, GSP schemes introduced after the second meeting of UNCTAD had many conditions that were not outlawed by the Enabling Clause, ^{Footnote 97} which means that the donors had never accepted that their freedom to impose conditions would be constrained. ^{Footnote 98} However, these arguments ignore the fact that the Enabling Clause failed to explicitly allow conditions in the GSP. In fact, the Appellate Body only interpreted paragraph 3(c) because it was already present in the Enabling Clause.

Other scholars argue that no conditions were envisaged when the GSP came about because it aimed at enhancing growth in developing countries by means of exports. ^{Footnote 99} Moreover, these conditions demonstrate the donors’ desire to maintain unilateralism in granting preferences, ^{Footnote 100} and amount to disguised protectionism. ^{Footnote 101} Importantly, these GSP schemes containing conditionalities demonstrate an element of discrimination and reciprocity not envisaged by UNCTAD Resolution 21(II). ^{Footnote 102} Most of the discriminatory characteristics in GSP schemes are not meant to promote social values and may be open to scrutiny following this case. ^{Footnote 103} In fact, the EU argued in this case that the drug arrangements were meant to protect European citizens from drugs, which proves that the conditions in GSP schemes may not necessarily be for the benefit of the beneficiaries.

Additionally, donors question the validity of the regulation of conditionalities in unilateral GSP schemes and might withdraw from these schemes if restrictions are imposed on their freedom to impose conditions. ^{Footnote 104} However, the lack of regulation of conditionalities might lead to discriminatory preferences, similar to those that existed before UNCTAD. ^{Footnote 105} It has been argued that developing countries should have anticipated these conditions because they help donors buy political support for the GSP schemes in their territories. ^{Footnote 106}

The EC – Tariff Preferences Case: An Alternative Approach

If the Appellate Body had been asked to decide on the WTO legality of conditionalities, it would have been in a position to say that conditions should not be attached to GSP schemes because conditions would necessarily be discriminatory. This would not have amounted to legislation because India had argued that all potential beneficiaries should be treated alike. There could be two responses to this argument. The first is that all potential beneficiaries should be required to fulfil the same conditions irrespective of their ability to do so (since similarly situated is not defined); this was the response given by the Appellate Body. The second is that no potential beneficiary should be required to fulfil any condition. In this second response, the Appellate Body could have compared paragraph 3(c) and footnote 3 of the Enabling Clause and could have concluded that modification of GSP schemes in accordance with paragraph 3(c) does not allow for conduct that would be considered discriminatory under footnote 3. Thus, conditions, which are discriminatory by nature, would have been outlawed. Conditions are discriminatory by nature because (1) they ignore the different situations of beneficiaries and (2) the power relationship between the donor and the beneficiary is asymmetrical. Thus, this response would confirm India’s argument that all potential beneficiaries should be treated alike. In addition, the removal of non-economic conditions in trade schemes would be helpful in achieving the economic development goals that were the initial reason for establishing the GSP.

Interpretations can also be suggested by applying the VCLT to interpret paragraph 3(c) and footnote 3 of the Enabling Clause. Article 3.2 of the Understanding on Rules and Procedures Governing the Settlement of Disputes (DSU) refers to the “customary rules of interpretation of public international law” to interpret WTO law. ^{Footnote 107} These customary rules have been codified in Articles 31 to 33 of the VCLT. The WTO panels and Appellate Body have always relied on these articles to interpret WTO law. Of course, the use of these interpretative aids can lead to different findings.

Article 31.1 of the VCLT states that “[a] treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.” The WTO panels and Appellate Body have shown great enthusiasm for using dictionaries to determine the ordinary meaning of treaty terms. ^{Footnote 108} But can any dictionary define “modified” in paragraph 3(c) and “non discriminatory” in footnote 3 of the Enabling Clause as allowing for conditionalities? The context of the terms of the treaty and the object and purpose of the treaty can be found in the preamble to the WTO Agreement, which states that

The Appellate Body did refer to developing countries at different levels of development, as mentioned in the preamble to the WTO Agreement, to interpret footnote 3 of the Enabling Clause. ^{Footnote 109} Conditionalities can be viewed as creating obstacles to, or encouraging, the achievement of the aforementioned goals. Beneficiaries need to be able to fulfil the conditionalities to benefit from these goals. Thus, conditionalities themselves are conditional on the capacity of the beneficiaries. In such a situation, the ordinary meaning of the terms “modified” and “non discriminatory” in their context and the object and purpose of the treaty would not allow for conditionalities.

The chapeau of Article 31.2 of the VCLT states that the context of a treaty includes its preamble and, thus, supports the aforementioned interpretation. Article 31.2(a) of the VCLT states that the context also includes “[a]ny agreement relating to the treaty which was made between all the parties in connexion with the conclusion of the treaty.” The panel had ruled that UNCTAD Resolution 21(II) was such an agreement under Article 31.2(a) and that it was context for interpreting the Enabling Clause. ^{Footnote 110} This resolution makes no mention of conditionalities. The Enabling Clause, when interpreted in the context of this resolution, would not allow for conditionalities. Article 31.2(b) of the VCLT states that this context also includes “[a]ny instrument which was made by one or more parties in connexion with the conclusion of the treaty and accepted by the other parties as an instrument related to the treaty.” UNCTAD Resolution 21(II) would be such an instrument, and it effectively disallows conditionalities under the Enabling Clause.

This interpretation is also supported by the Appellate Body’s ruling in European Communities – Customs Classification of Frozen Boneless Chicken Cuts (EC – Chicken Cuts), in which the Appellate Body stated that the concept of context under Article 31 is not limited to the treaty text as demonstrated by Articles 31.2(a) and 31.2(b). ^{Footnote 111} Thus, the context for interpreting the WTO Agreement (to which the GATT 1994, along with the Enabling Clause, is annexed) can be found outside it as well. The Appellate Body made this observation while stating that the Harmonized Commodity Description and Coding System (Harmonized System), which is not part of WTO law, constituted context under Article 31.2(a) for the interpretation of the schedules of members, ^{Footnote 112} which are WTO law as per Article II.7 of the GATT. ^{Footnote 113} The Appellate Body arrived at this conclusion in view of the close connection between the Harmonized System and WTO law. ^{Footnote 114} Applying this logic, there is a close connection between UNCTAD Resolution 21(II) and the 1971 GSP Waiver, which is also referred to in footnote 3 of the Enabling Clause. Thus, this resolution can be context for the interpretation of the Enabling Clause, as stated by the panel.

Article 32 of the VCLT states that the supplementary means of interpretation include “the preparatory work of the treaty and the circumstances of its conclusion.” The panel ruled that UNCTAD Resolution 21(II) was preparatory work to interpret the Enabling Clause. ^{Footnote 115} This would disallow conditionalities. In any case, Article 32 provides a non-exhaustive list of the supplementary means of interpretation, as highlighted by the Appellate Body in EC – Chicken Cuts. ^{Footnote 116} Therefore, the panel was only doing what was allowed in the VCLT. As stated earlier, these articles can lead to different findings, and it is possible to argue that the preamble to the WTO Agreement would allow conditionalities. Moreover, it can also be argued that UNCTAD Resolution 21(II) does not outlaw conditionalities and, therefore, that the Enabling Clause can be interpreted to allow them. However, the aim of the UNCTAD negotiations on GSP was to reduce discrimination in the existing preferences, which was incorporated in the 1971 GSP Waiver and the Enabling Clause. Yet the Enabling Clause did not outlaw conditions that can lead to discrimination. ^{Footnote 117} But it did not allow conditions either, and this should be taken into account in a judicial proceeding.

Apart from these articles, the preamble of the VCLT mentions certain principles that have not been used by the WTO panels and Appellate Body in interpreting WTO law. These principles include the following:

The question is whether these principles can be used by the WTO panels and Appellate Body to interpret the Enabling Clause. WTO law is part of public international law even in the absence of Article 3.2 of the DSU ^{Footnote 118} and “is not to be read in clinical isolation from public international law.” ^{Footnote 119} In addition, Article 31.3(c) of the VCLT states that “[a]ny relevant rules of international law applicable in the relations between the parties” must be taken into account. Of course, WTO law would prevail over public international law in case of conflict between the two. ^{Footnote 120} It has been argued that the principles mentioned in the preamble of the VCLT are constitutional principles and have been enshrined in many instruments of the United Nations (UN), thus making them customary law requirements for the members of the UN and for the interpretation of treaties, including in WTO judicial proceedings. ^{Footnote 121} Article 31.3(c) of the VCLT reflects the acceptance of these principles by all members of the UN. ^{Footnote 122} In view of the fact that most members of the WTO are also members of the UN, this would imply that WTO law can be interpreted in light of the principles mentioned in the preamble of the VCLT.

In any case, the Appellate Body has inherent powers to administer justice, but it has not exercised these powers. ^{Footnote 123} And how would justice be defined? It could mean that conditions in GSP schemes should be imposed to benefit citizens of beneficiary countries. It could also mean that imposing conditions would impose unjust burdens on the beneficiary countries. In addition, not giving additional preferences or taking away preferences when conditions are not fulfilled means that the citizens of beneficiary countries are being deprived of employment opportunities. Moreover, a balance will have to be struck between the rights of states (sovereign equality, independence, and non-interference) and the human rights of citizens. This is because many developing countries may not be parties to many international instruments. ^{Footnote 124} Even if the rights of states can be ignored, the Enabling Clause should be interpreted to allow conditionalities only if they promote human rights and other associated rights. This would also take into account the meaning of the term “justice.” Furthermore, top-down approaches are not very effective in achieving their goals, and bottom-up approaches pushed by citizens are preferable. ^{Footnote 125} This would mean that conditionalities may not achieve their goals even if they are well intentioned unless the citizens of beneficiary countries push for change.

In addition, the WTO panels and Appellate Body can come up with a set of rules to examine legitimate conditions in GSP schemes and resolve any disputes about them. This allows GSP donors to determine conditions unilaterally, subject to the justification of differential treatment of the beneficiaries and judicial review. However, the GSP donors may design schemes that fulfil their policy goals and comply with the Appellate Body’s requirements. Or they might have many conditions in their GSP schemes, thus shifting benefits from one beneficiary to another. The beneficiaries may not challenge these conditions for the usual reasons of resource insufficiency or the political weight of the GSP donors as well as because they would be unsure of the result of the challenge due to the uncertainty created by the Appellate Body’s decision. ^{Footnote 126}

These were suggested interpretations. In view of the existing interpretation given by the Appellate Body, the EU replaced the defaulting regulation with another one.

Evaluation of the New European GSP + Scheme

The EU came up with Council Regulation (EC) 980/2005 Applying a Scheme of Generalised Tariff Preferences ^{Footnote 127} to replace the previous Regulation 2501/2001. This was followed by Council Regulation (EC) 732/2008 Applying a Scheme of Generalised Tariff Preferences. ^{Footnote 128} Article 1.2 of the two regulations laid down the following three schemes:

• • a general arrangement;

• • a special incentive arrangement for sustainable development and good governance; and

• • a special arrangement for least developed countries.

Article 1.2 of the currently applicable Regulation (EU) 978/2012 Applying a Scheme of Generalised Tariff Preferences (Regulation 978/2012) provides for the following schemes:

• • a general arrangement;

• • a special incentive arrangement for sustainable development and good governance (GSP +); and

• • a special arrangement for the least-developed countries (everything but arms (EBA)). ^{Footnote 129}

The European Economic and Social Committee (EESC) has stated that the special incentive arrangements in Regulation 2501/2001 for the protection of labour rights and of the environment have not been effective and should be withdrawn. It has also stated that conditionalities will not be useful in achieving their aim unless sufficient preferences are granted. ^{Footnote 130} Despite this statement, the EU came up with the aforementioned three regulations containing a special incentive arrangement for sustainable development and good governance because the EU felt that it had absolute discretion to impose conditions on preferential trade. ^{Footnote 131} In fact, the EU stated that it was clear that it should promote human rights, labour rights, the environment, and good governance standards by means of the GSP +. ^{Footnote 132} Apart from the GSP +, Regulation 978/2012 also lays down the following vulnerability criteria:

• • the seven largest GSP sections of a beneficiary’s exports to the EU must exceed 75 percent of its total exports to the EU during the last three consecutive years and

• • a beneficiary’s exports of particular products to the EU must not exceed 2 percent of the EU’s imports of those products during the last three consecutive years. ^{Footnote 133}

This does not fulfil the Appellate Body’s criteria of differentiating between beneficiaries on the basis of development, finance, and trade needs. ^{Footnote 134} Thus, it is obvious that there is a lack of coherence and a contradiction in the trade and development policy of the EU. ^{Footnote 135}

Despite various regulations coming and going, the special incentive arrangement for sustainable development and good governance known as the GSP + has remained the same. ^{Footnote 136} It aims to help countries most in need. ^{Footnote 137} Developing countries wanting to benefit from it must fulfil the vulnerability criteria as well as ratify and effectively implement twenty-seven international conventions on human rights, labour, environmental protection, the fight against drugs, and corruption. ^{Footnote 138} Regulation 2501/2001 only referred to conventions on labour and the sustainable management of tropical forests. ^{Footnote 139} Thus, the potential beneficiaries could choose from any of the three arrangements devoted to labour, environment, and drugs and fulfil only the requirements pertaining to that particular arrangement. In the GSP +, a beneficiary has to fulfil all of the criteria. So the GSP +, which was introduced after the EC – Tariff Preferences case, requires the fulfilment of more conditions. Therefore, the GSP + is more restrictive and could even be called protectionist. ^{Footnote 140} It is doubtful that many developing countries can fulfil these conditions. ^{Footnote 141} In addition, the list of conventions to be ratified and effectively implemented is a unilateral decision taken by the EU. How does the Appellate Body propose to regulate the objectivity or subjectivity in this decision? It is possible that the conventions are chosen to benefit certain countries. This shows that the Appellate Body’s dependence on international instruments is not all that helpful in eliminating discrimination.

On the face of it, this scheme is not discriminatory and opaque since all developing countries are eligible for preferences if they fulfil the criteria that are laid down. But all developing countries may not be in a position to fulfil all of the criteria. Even the EESC has recognized that the fulfilment of conditions entails a high cost that is not necessarily offset by preferential treatment. ^{Footnote 142} So a de jure non-discriminatory scheme may lead to de facto discrimination. Countries may have the same needs irrespective of their capacity to ratify and implement these conventions. Or countries may have different needs but are required to ratify and implement the same conventions. Thus, it is arguable if the GSP + entitles similarly situated beneficiaries to similar preferences.

In any case, beneficiaries are not necessarily similarly situated because they ratify and implement the same conventions. ^{Footnote 143} Each developing country may have different needs that are not necessarily reflected in multilateral instruments, especially those included in the GSP +. Thus, the GSP + may not respond to the Appellate Body’s criterion that the preferences alleviate a need. This means that the Appellate Body’s and EU’s standard of defining a need in terms of ratification and implementation of conventions does not necessarily fulfil the aim of addressing the need through preferences, thus defeating their very purpose. According to the European Commission, granting additional preferences depends on an objective identification of the development needs of developing countries. ^{Footnote 144} Are conditionalities requiring the ratification and implementation of international conventions an objective method of identifying a development, financial, or trade need?

According to the European Parliament, these requirements can create barriers to the trade of developing countries, thus excluding them from potential preferences. ^{Footnote 145} Therefore, discrimination is still possible if the standards used to determine the eligibility of the beneficiaries are not objective or are unjustifiably burdensome. ^{Footnote 146} As much as a single arrangement might seem simpler, it requires the fulfilment of extensive criteria defined in terms of international conventions and agencies, which might make it onerous ^{Footnote 147} for potential beneficiaries. Furthermore, a really backward country would find the fulfilment of these criteria unjustifiably burdensome compared with a more advanced country. ^{Footnote 148} This shows that standardized conditionality for beneficiaries at different levels of development can lead to discrimination, against which the criteria laid down by the Appellate Body in the EC – Tariff Preferences case does not provide a guarantee.

In United States – Import Prohibition of Certain Shrimp and Shrimp Products, the Appellate Body stated that “[w]e believe that discrimination results not only when countries in which the same conditions prevail are differently treated, but also when the application of the measure at issue does not allow for any inquiry into the appropriateness of the regulatory program for the conditions prevailing in those exporting countries.” ^{Footnote 149} In other words, discrimination will result since the GSP + requires potential beneficiaries to fulfil identical conditions irrespective of their different situations/capabilities. Furthermore, WTO law itself shows the way forward insofar as Article XVII.3 of the General Agreement on Trade in Services on national treatment states that “[f]ormally identical or formally different treatment shall be considered to be less favourable if it modifies the conditions of competition in favour of services or service suppliers of the Member compared to like services or service suppliers of any other Member.” ^{Footnote 150} Thus, there is a recognition that formally identical treatment can lead to discrimination. In this sense, the complete absence of conditions from GSP schemes is identical treatment, but it is less problematic than having conditions.

Another problematic issue is that the potential beneficiaries are required to fulfil the conditions before benefiting from the preferences. ^{Footnote 151} This discriminates in favour of those potential beneficiaries that can incur the cost of fulfilling the conditions. This shortcoming may be removed by (1) granting preferences before requiring fulfilment of the conditions and by (2) introducing a time gap between granting the preferences and fulfilling the conditions. Of course, this issue is not problematic if the conditions reflect customary international law. ^{Footnote 152} Additionally, the GSP + uses the conclusions of the monitoring bodies under the relevant conventions to determine if potential beneficiaries have effectively implemented the conventions and if they are eligible to receive preferences under the GSP +. ^{Footnote 153} However, the Appellate Body’s criteria did not account for the lack of clarity in the donors’ standard of review applicable to the conclusions of the monitoring bodies under the relevant conventions to determine if potential beneficiaries have effectively implemented the conventions. For example, under the GSP +, the type of infraction of the convention(s) leading to ineligibility of preferences is not known. ^{Footnote 154}

So is defining conditionalities in a way that does not allow all potential beneficiaries to fulfil them not a form of discrimination? This method of defining conditionalities is legal because it fulfils the criteria laid down by the Appellate Body. But the Appellate Body’s criteria are flawed and could lead to a return of the discriminatory preferences that existed before UNCTAD. ^{Footnote 155} The other question is whether preferential trade based on conditionalities actually responds to a development need that is not entirely economic in nature. For example, can preferential trade prevent genocide, in view of the requirement in the current GSP + to ratify and implement the requisite convention? ^{Footnote 156} Thus, the GSP + may be open to challenge if it does not alleviate the needs for which it was enacted. ^{Footnote 157}

According to the European Commission, there is a link between development and human rights, labour rights, environmental protection, and good governance. Additional preferences can respond positively to these specific needs. ^{Footnote 158} However, the EESC has stated that “there is no evidence that the special incentive arrangements for combating the production and trafficking of drugs, from which twelve countries have benefited, has had any impact whatsoever on the drug trade.” ^{Footnote 159} It has also stated that the GSP aims to develop the economies of beneficiaries instead of resolving all of their problems. ^{Footnote 160} This lack of consensus within the EU supports the argument of beneficiaries that these conditions are foreign to the preferences devoted entirely to trade policy. ^{Footnote 161} In addition, granting additional preferences on the fulfilment of conditions implies granting preferences to countries that may not necessarily be in need of help and depriving those countries that cannot fulfil these conditions but that may really be in need of help. ^{Footnote 162} These countries may then question how these conditionalities respond to development needs. In any case, the success of the GSP + depends on the meaningful selection of beneficiaries, contingent on the use of relevant international conventions and their monitoring and review mechanisms. ^{Footnote 163} In addition, potential beneficiaries’ acceptance of conventions does not necessarily imply the acceptance of the use of the conventions as conditionalities leading to discriminatory trade treatment. ^{Footnote 164} Moreover, an important aspect of the GSP + is that it does not benefit countries that have implemented the content of these international conventions without ratifying them.

The debate between the donors and the beneficiaries can go on endlessly, but there is some merit in the argument that the norms enshrined in international conventions can lead to development. The real question is whether or not conditions are the appropriate means of achieving this aim. Regarding the conditionalities of labour standards, certain scholars argue that trade liberalization will lead to an improvement in the condition of labour, whereas other scholars argue that explicit linkages between trade policy and labour standards are required. ^{Footnote 165} Others feel that the imposition of higher standards will lead to higher costs of hiring labour, thus leading to the employment of fewer workers. Consequently, higher standards will only benefit the fewer workers employed and not the large number of unemployed persons. ^{Footnote 166}

Therefore, despite the benefits of labour standards, the linkages between trade policy and labour standards should be approached cautiously, keeping in mind the situation of employed and unemployed persons. ^{Footnote 167} A lot has been written on the issue, but there is no consensus on the utility of trade conditionality in promoting human and labour rights. ^{Footnote 168} There are concerns that labour conditionalities are a result of protectionist interests in the developed world. ^{Footnote 169} The withdrawal of such preferences will certainly have a negative effect on the populations of beneficiaries working in the export sectors, ^{Footnote 170} which means that conditionalities have been imposed without proper research on their efficacy. However, such research will only help in devising more effective conditionalities, which will be helpful for donors and beneficiaries.

What Can GSP Donors Learn from the Discourse on Conditionalities in the Loans Granted by the World Bank?

Conditionalities are not unique to the GSP. ^{Footnote 171} They are used in various international transactions. For example, conditionalities are also part of the loans granted by the World Bank. The issues that arise in the case of the GSP + have also arisen in the case of the World Bank. Suggestions have been made to resolve the problems that arise in the case of conditionalities imposed by the World Bank. These suggestions may provide some guidance in resolving the problems that arise in the case of conditionalities imposed in the GSP +.

In the World Bank, the impact of conditionalities on the performance of the recipient country has been mixed, tending towards the negative. ^{Footnote 172} Moreover, the number of conditions imposed has been very high. ^{Footnote 173} This is also the case with the EU GSP +, which requires the ratification and effective implementation of twenty-seven conventions. Therefore, donors should limit the lending conditions, which would actually help recipient governments focus on the limited conditions they can fulfil rather than trying to fulfil unimportant ones. Recipient countries have limited resources, and the imposition of too many conditions is burdensome.

Questions have also been raised on the imposition of conditions by donors, thus undermining the idea of recipient countries having ownership over their agenda on good governance. In fact, the major criticism of conditionality is its dictatorial or coercive element in imposing policies that makes for a weak partnership between donors and recipients because it deprives the latter of ownership over their choice of policies. ^{Footnote 174} In addition, there is no one single policy reform appropriate for all recipients. Thus, conditionality should differ according to the country context ^{Footnote 175} since standardized conditionality does not take into account the situation prevailing in different countries and may lead to recipient countries losing ownership of the process. Moreover, studies prove that standardized conditionality in countries at different levels of development does not work. Recipient countries may not be able to fulfil the conditions, and, thus, they would lose the aid. However, it is possible to work with donors to reduce the number of conditionalities and to make them more specific. It is also possible to moderate the conditionalities according to the situation — for example, conditionalities could be different in a crisis situation. ^{Footnote 176} Additionally, the recipient country’s institutional capacity should be taken into account since it determines the success of the conditionalities. ^{Footnote 177} The generality of these issues was reflected in a specific study on Vietnam, which concluded that externally imposed conditionalities are not helpful. Therefore, an open dialogue between the donor and the recipient country is required, which should lead to the recipient country taking full ownership of the design and implementation of its commitments. ^{Footnote 178} Such an open or policy dialogue is like a partnership in which donors persuade or convince recipient countries instead of coercing them to adopt certain policies. Such a dialogue is more likely to be successful, and conditions can be used to support this dialogue. ^{Footnote 179} These suggestions could be helpful for conditionalities in GSP schemes since they allow for a dialogue between the GSP donor and the potential beneficiaries. However, the dialogue should not degenerate into a mechanism to pressure the potential beneficiaries to comply with the conditions or to benefit the donors’ allies.

The fact remains that conditions may not be entirely effective, whether used independently or with dialogue. This is because governments are reluctant to implement policy reforms if the conditions are too restrictive, and conditions are obviously inappropriate in cases where governments are keen on policy reform. ^{Footnote 180} Furthermore, reform is a slow process, and conditionalities try to impose a quick reform, while ignoring the local conditions, thus leading to failure. Consequently, donors should support policy reform instead of imposing it. ^{Footnote 181} They can influence the policy reform in recipient countries by providing analytical research on the effects of alternative policies. Donor influence undermines the ownership of policy reform, but it is acceptable as long as the choice of policy reform, with the belief that it is the best choice, originates in the recipient country’s government. Thus, donors can provide information on policies, but the choice should be left with the recipient country’s government. Additionally, donors should not provide standardized advice to all recipient governments unless it is the correct advice for that country. This allows recipient governments to experiment with policies and does away with the coercive nature of conditionalities since influence is not equivalent to the imposition of policies.

The other option is for recipient governments to do their own research and come up with their choice of policy, which would truly amount to ownership of the policy reform process. However, this is not necessarily required for policy reform. In addition, recipient countries’ governments may be willing, but unable, to implement the chosen policies. In such a situation, donors should help in the implementation of the policies by providing technical assistance. ^{Footnote 182} In fact, the EU has been funding a project since 1 October 2015 in four of its GSP + beneficiaries to help them develop their capacity to comply with labour standards. ^{Footnote 183} Going further, donors should grant benefits even if the policy reform does not materialize. ^{Footnote 184} As opposed to conditionalities, such a dialogue would be slow in bringing about reform but would be adapted to the recipient country’s environment. ^{Footnote 185}

As stated earlier, conditions can be used to support dialogue. These conditions should be precise, clear, easy to monitor, and should reinforce or ensure that the recipient governments actually follow the policy choices they have made, instead of having policies imposed on them. Also, conditions should be defined in terms of policy choices and not in terms of the goal the policy choice aims to achieve because goals may not be achieved even if the policy reform has been properly implemented. In addition, conditions in various policy areas should be independent of each other to avoid conflict between them. However, conditionalities in the traditional sense are not the best means of policy reform since they do not bring about change in the beliefs of the recipient governments. Change in beliefs can be brought about by starting with simpler reforms, keeping in mind that the breadth of the reform program is directly proportional to the level of development in the recipient country. It is not possible to have a dialogue with failed states or authoritarian regimes. However, these are exceptions and do not reflect a shortcoming of the policy dialogue since traditional conditionalities would also be ineffective in such a situation. ^{Footnote 186}

Even though ownership is preferable to conditionalities, the issue of power remains unchanged. Thus, real partnerships should include shared objectives, mutual accountability, and the two-way exchange of knowledge. However, these may not materialize if the balance of power is tilted in favour of the donor. Since the donors provide the benefits, they expect accountability from the recipient countries. The governments of recipient countries should also be accountable to their populations. ^{Footnote 187} This is a significant issue in terms of conditionalities, ownership, partnership, or any kind of benefit because, ultimately, it is the population of the recipient country that must benefit from this process. Moreover, the presumption that donors have the knowledge on policies means that the exchange of knowledge is not a two-way process. ^{Footnote 188}

The same issue has been highlighted in the case of the GSP + conditionalities. The GSP donor makes an independent decision on the conditions to be fulfilled by the beneficiaries. These conditions may involve the ratification of various treaties and so on, but they certainly ignore the knowledge of the beneficiaries regarding their choices and needs, ^{Footnote 189} despite the EESC suggesting that consultations should be held with stakeholders in developing countries before the introduction of a new GSP scheme. ^{Footnote 190} Moreover, the ownership model is persuasion based and sees benefits as steps to inducing reforms in the recipient country. This shows that it is a paternalistic model as opposed to a partnership model. ^{Footnote 191} The same problem arises in the case of conditionalities in the GSP + since they do not involve the concept of ownership, let alone the concept of partnership.

However, this paternalism may not necessarily encourage the desired reforms, as seen in the case of the drug arrangements. A real partnership would require that donors remove reform-inducing conditions and standardized conditionality, allowing for more diversity in the actual reforms. Donors should also be held accountable, especially to check if the benefits they are granting are being diluted in other sectors ^{Footnote 192} and if the implementation of the reforms they are suggesting is leading to any problems in the recipient country. ^{Footnote 193} The same issues arise in the case of conditionalities in the GSP + where the EU has decided on a list of conventions to be ratified irrespective of the country that applies for the additional preferences and where there is no scrutiny to find out if the preferences granted are being weakened by other aspects of EU policy, including trade policy. This situation is happening despite the fact that civil society representatives have stated before the EESC that the benefits of the GSP are outweighed by non-tariff barriers. ^{Footnote 194} This stance against conditionalities recognizes that donors would like to have some conditions regarding the use of the benefits provided to the beneficiaries, but it does not support conditionalities that impose policy reform. ^{Footnote 195}