African Continental Free Trade Agreement's Conditional Most Favoured Nation: A Necessary Compromise?

Abstract

The inception of the African Continental Free Trade Agreement (AfCFTA) constitutes a major advancement in Africa's economic integration process. Diverging from what appears to be the norm in contemporary trade treaties, the agreement adopts a conditional Most Favoured Nation (MFN) clause hinged on the principle of reciprocity. Without the promise to reciprocate preferential treatment, the beneficiary state does not assume the right to demand MFN treatment. In broader discussions, this feature has been criticized for possessing the tendency to restrict trade. However, examining it in the context of Africa's trade paradigm, this paper argues that the non-automaticity of the AfCFTA's MFN clause is a cardinal feature safeguarding its existence.

1. Introduction

The Most Favoured Nation (MFN) clause has the potential to convey unintended consequences in international commercial relations.¹ To this extent, drafters of MFN clauses are caught up in the dilemma of either drafting them in general terms, thereby risking their effectiveness through interpretation of the ejusdem generis rule or drafting them in an explicit manner with an enlisted scope that also avails the prospect of incomplete enumeration.² Unless the clause is restricted to issues where there is significant identity between the subject matter of the two concerned clauses, the conclusion of a case is likely to be an imposition on the granting state obligations it had never contemplated.³ For this reason, many drafters of international agreements choose to specify the scope and exceptions of contemporary MFN clauses. Such precisions and peculiarities invoke unique implications in international law.

In line with the preamble of the World Trade Organization's (WTO) General Agreement on Tariffs and Trade (GATT), the Agreement Establishing the African Continental Free Trade Area (AfCFTA Agreement) embodies reciprocal and mutually advantageous arrangements.⁴ The AfCFTA is an ambitious effort toward Africa's economic integration and a positive contribution to global trade liberalization. As with the Boosting Intra-African Trade (BIAT) initiative, the AfCFTA seeks to enhance intra-African trade, which constitutes about 16% of the continent's trade.⁵ It is a comprehensive framework that covers the areas of trade in goods, investment, trade in services, competition policy, and intellectual property rights.⁶ Among its governing principles are national treatment, variable geometry, the acquis, reciprocity, and MFN treatment.⁷

Though highly inspired by the GATT, the architecture of the MFN provisions featured in the AfCFTA Protocol on Trade in Goods (PTG) appear uniquely crafted to suit specific purposes. With the MFN clause being noted as a pillar towards trade liberalization, the widely held notion is that any derail in its effectiveness is a hindrance to international trade.⁸ Departing from what has widely been observed as a traditional MFN clause (unconditional MFN), the AfCFTA's PTG MFN clause stipulates a reciprocal condition upon which it may be triggered.⁹ Some have labelled this form technically unworthy to constitute an MFN clause¹⁰ and, hence, of little practical value.¹¹ Others have argued that the guarantee offered by an unconditional MFN clause is more valuable than any guarantee that could be received under a reciprocal MFN clause.¹² It is therefore upheld that the reciprocal MFN clause constrains international trade.¹³

In contrast with this perspective, this paper argues that the reciprocal MFN clause featured in the AfCFTA PTG is a necessary compromise. In doing so, this piece: (i) unpacks the MFN clause in the AfCFTA PTG and establishes that it is a conditional MFN clause; (ii) argues that the non-automaticity of the AfCFTA's MNF clause is a cardinal feature safeguarding the AfCFTA's existence.

2. Background of Conditional MFN

Besides minor variations, the MFN clause mainly appears in two forms: conditional and unconditional MFN clauses. Until the year 1778, when the conditional form of the clause first appeared in a commercial treaty between the United States of America (USA) and France, the MFN clause was widely observed in its unconditional form where there was no obligational limitation to contracting parties.¹⁴ Article II of the treaty mentioned that:

The Most Christian King and the United States engage mutually not to grant any particular favour to other nations, in respect of commerce and navigation, which shall not immediately become common to the other Party, who shall enjoy the same favour, freely, if the concession was freely made, or on allowing the same compensation, if the favour was conditional.¹⁵

Since then, the USA has, through most of its history, been at the forefront of advocacy in favour of the conditional MFN.¹⁶ This was within the mercantilist period where the pursuit of ‘stateness’ was the order of the day,¹⁷ and the concept of international trade at the time was to maximize exports while restricting imports. Trade was bilateral centered,¹⁸ and the conditional MFN was used ‘to safeguard preferential treatments accorded in a bilateral relationship’.¹⁹ It is believed that the conditional clause was mainly adopted by the USA, a net importer, to protect its emerging industries.²⁰ This advantage of the clause gives a glimpse of its essence in the AfCFTA, considering Africa's continuous reliance on imports and its low industrialization rate. Conditional MFN clauses were also frequently used among European nations between 1820 and 1860 and widely utilized by Japan and Latin-American states.²¹ Bailey stresses that appetite for the conditional clause stemmed from the loss of confidence in the unconditional one.²² Although it was argued that the clause was made conditional upon the insistence of France, conditional MFN clauses for a long time suited the USA's political and economic interests.²³ The phrase ‘freely, if the concession was freely made, or on allowing the same compensation, if the favour was conditional’ became the standard model for all USA commercial treaties until 1923,²⁴ when in an agreement with Brazil, the unconditional MFN was adopted.²⁵ Subsequently, the unconditional form of the clause has dominated commercial treaties.

The rationale behind the conditional clause can be deduced from a report delivered to the USA's Congress by John Jay, the then Secretary for the Department of Foreign Affairs, on the rightful interpretation of the MFN clause in the Treaty of 1782 with the Netherlands. The report stated that:

it would certainly be inconsistent with the most obvious principles of justice and fair construction, that because France purchases, at a great price, a privilege of the United States, that therefore the Dutch shall immediately insist, not on having the like privileges at the like price, but without any price at all.²⁶

The reasoning here is that such an arrangement (the unconditional MFN) would be unjust not only to the USA but also to France. This dictum of fairness has popularly been expressed as a major advantage to the conditional clause.²⁷ The reciprocal MFN is also noted to further strengthen unity among parties involved due to its consultation-driven approach.²⁸ In this sense, for the clause to thrive, there should be already established political ties and cultural coordination among concerned state parties.²⁹ This aligns with the African Union's (AU) vision of utilizing the AfCFTA as a stepping stone to its Agenda 2063³⁰ and further justifies its adoption in the AfCFTA Agreement, noting the already existing political and cultural cohesion among African states.

3. MFN Scope in Trade

The MFN clause has been described as the cardinal organizing provision of the GATT and the global trading system of rules.³¹ Its inception can be traced as far back as the 11th century.³² Since the twentieth century, however, the MFN principle has remained a key feature of international commercial agreements³³ to the extent that the International Law Commission has established Draft Articles on Most-Favoured-Nation Clauses (ILC Draft).³⁴ Inferring from the fact that it has never been challenged and the frequency of its application in defining MFN clauses and determining their scope and application, the ILC Draft remains an authoritative source in international law concerning the interpretation and application of MFN clauses.³⁵ Though non-binding, the ILC Draft Articles provide the legal structure and codification for drafting MFN clauses in international treaties. Per its definition, an MFN clause is ‘a treaty provision whereby a state undertakes an obligation towards another state to accord most-favoured-nation treatment in an agreed sphere of relations’.³⁶ In short, the MFN clause stipulates the agreed MFN treatment in an international agreement. The MFN treatment per Article 5 of the ILC Draft is a:

treatment accorded by the granting State to the beneficiary State, or to persons or things in a determined relationship with that State, not less favourable than treatment extended by the granting State to a third State or to persons or things in the same relationship with that third State.³⁷

The treatment mitigates discrimination and tests the limits of international law by breaking treaty borders.³⁸ While MFN clauses protect contracting parties against discrimination, their advantages transcend beyond a normal clause promising non-discrimination. The MFN clause promises that state parties to a treaty will offer each other the best treatment they provide to third-party states. Thus, MFN clauses generally import favourable treatments from external treaties of the granting state and, in that sense, only become useful when the granting state offers more favourable treatment to a third-party state.³⁹ Its application is so critical that, unless unequivocally worded, an MFN clause is deemed to affect both past and future third-party agreements.⁴⁰ This implies that unless an MFN clause states explicitly that it is not to be applied to treaties signed before the inception of the treaty containing the clause, it is deemed applicable to both past and future treaties of the parties involved. This is rationalized in the sense that since the motive of the MFN clause is to put parties to a treaty and related third parties on equal footing, its applicability to both past and future agreements, unless expressly worded, is an act of good faith.⁴¹

Besides the guidelines provided by the ILC Draft, there is no legal manual or restraint to the drafting of MFN clauses and, as a result, they differ in length and depth. In practice, MFN clauses are framed to reflect the underlying objectives of an international agreement. They take varied forms in different treaties.⁴² Historically, however, the predominant vagueness of MFN clauses has triggered legal disputes and unpredictability of dispute outcomes.⁴³ Although there has been dormancy (with regard to trade) in MFN related disputes in recent times, historical happenings express the critical implication of MFN wording.⁴⁴

Article 9:1 of the ILC Draft confines the scope of MFN clauses to ‘only those rights which fall within the limits of the subject-matter of the clause’⁴⁵ in relation to ‘persons or things which are specified in the clause or implied from its subject-matter’.⁴⁶ Therefore the MFN clause can be triggered in relation to rights conferred by other treaties ‘in regard to the same matter or class of matter’.⁴⁷ This rule is generally termed the ejusdem generis rule.⁴⁸ It rests on the common assumption that the clause should be restricted within the scope of the subject matter intended by the parties when inserting it into the treaty.⁴⁹ Therefore, the right to demand favourable treatment under the MFN clause ‘only’ arises when the granting state accords favourable treatment ‘within the limits of the subject-matter of the clause’ to a third-party state.⁵⁰ In simple terms, the general principle is that states cannot be bound beyond the obligations they undertook.⁵¹ This connotes that although the MFN clause breaks treaty barriers, its scope of application is restrained to the subject matter of the clause in the base treaty and therefore establishes the confines within which the discussions in this paper are made.

4. Most Favoured Nation Principle under AfCFTA Trade in Goods

This section establishes that the AfCFTA PTG entails a reciprocal MFN clause and is therefore distinctive from that of the GATT, the Economic Partnership Agreement between West Africa and the European Union (EU-ECOWAS),⁵² and the Economic Partnership Agreement between the East African Community and the European Union (EU-EAC).⁵³ It thoroughly analyzes the wording of the PTG MFN provisions and examines its applicability in practical terms. We make some comparisons with the GATT, EU-ECOWAS, and EU-EAC and examine the different approaches the PTG MFN clause adopts in relation to pre- and post-AfCFTA agreements.

4.1 AfCFTA in a Nutshell

The AfCFTA constitutes a continent-wide free trade area that merges Africa into a single market. Out of 55 African states, Eritrea remains the only non-state party to the agreement. It makes up the world's largest free trade area in terms of membership size and forms a major component of the AU's Agenda 2063.⁵⁴ With a population of 1.3 billion and a combined Gross Domestic Product (GDP) of US$3.4 trillion, the agreement is expected to rescue 30 million people from extreme poverty and ensure a US$450 billion boost in Africa's income by 2035.⁵⁵ Simo considers it the ‘solution to Africa's economic take-off’⁵⁶ while Magwape terms it Africa's ‘new hope’.⁵⁷ The agreement's main objective is to trigger a major boost in intra-African trade. This is to be achieved by eliminating barriers to trade in goods and a progressive liberalization of trade in service.⁵⁸ Per Songwe's assessment, it is the ‘game changer’ for intra-African trade.⁵⁹

4.2 Preferential Treatment

Among the guiding principles of the AfCFTA is the MFN principle. MFN right under the AfCFTA PTG emerges from Article 4:1 which provides that: ‘State Parties shall accord most-favoured-nation Treatment to one another in accordance with Article 18 of the Agreement’.⁶⁰ Article 18:1 of the AfCFTA Agreement in reference bases its MFN treatment on ‘preferences’ in general, whilst the GATT I:1 and EU-ECOWAS 16:2 specify ‘advantage, favour, privilege or immunity’ and ‘more favourable tariff treatment’ respectively. Like the AfCFTA PTG, the EU-EAC 15:1 also generally refers to ‘more favourable treatment applicable’. However, Article 4:2 of the PTG narrows the said ‘preferences’ down to ‘any advantage, concession or privilege granted’. It should be noted that the term ‘advantage’ constitutes a broad interpretation as provided by the panel in the United States–Non-Rubber Footwear.⁶¹

Again, the framing of the PTG MFN provisions does not provide any specificity to the scope of trade activities to which the ‘advantage, concession or privilege granted’ are applicable. Similarly, no such specification was made in the EU-EAC as Article 15:1 and 2 rather stress its generality with the phrase ‘with respect to the goods covered by this Part’. In comparison with the WTO legal framework, Article I:1 of GATT specifically limits the scope of trade activities concerned to: ‘customs duties and charges … method of levying such duties and charges … all rules and formalities in connection with importation and exportation … all matters referred to in paragraphs 2 and 4 of Article III … ’.⁶² Also, the EU-ECOWAS, by use of the term ‘more favourable tariff treatment’, clearly defines the limits of its MFN applicable activity to tariffs.⁶³ Without such specification, the PTG professes a contention of broader interpretation. It can therefore be established that the PTG MFN provisions, like the EU-EAC, are more generally worded and may imply broader application. That notwithstanding, without substantive case law and express legal provisions, it is difficult to comprehensively determine the spectrum of preferential treatments covered by the PTG MFN clause.

4.3 Conditionality

Stipulating the guidelines for MFN application, Article 18:1 of the AfCFTA Agreement provides that: ‘Following the entry into force of this Agreement, State Parties shall, when implementing this Agreement, accord each other, on a reciprocal basis, preferences that are no less favourable than those given to Third Parties.’⁶⁴ The ILC Draft emphasizes that an MFN clause can be made conditional to reciprocity by the wording of the clause.⁶⁵ Therefore, note should be taken of the term ‘on a reciprocal basis’ as required by Article 18:1 of the agreement. The term conditions the PTG MFN clause by making its extension dependent on a reciprocal exchange of concession.⁶⁶ Though MFN clauses are in principle, generally reciprocal in nature, this form of reciprocity is conditional.⁶⁷ It is a simplified version of America's ‘freely, if the concession was freely made, or on allowing the same compensation, if the favour was conditional’.⁶⁸ Besides Articles 18:3 of the AfCFTA Agreement and 4:1 of the PTG, all the AfCFTA MFN provisions (in relation to trade in goods) make use of the term ‘on a reciprocal basis’. Article 4:1 which excludes the term stipulates that the MFN treatment should be accorded ‘in accordance with Article 18 of the Agreement’. As earlier mentioned, all provisions under Article 18 of the AfCFTA Agreement with the exception of 18:3 possess the term ‘on a reciprocal basis’. It should also be noted that Article 18:3 of the agreement does not stipulate how MFN should be accorded but only guarantees the continued existence of preexisting third-party agreements. Hence, the PTG adopts a conditional MFN clause subject to reciprocal treatment.

MFN rights under this clause arise when preferential treatment is offered by the granting state to a third-party state and the agreed reciprocal treatment is accorded by the beneficiary state to the granting state.⁶⁹ It is terminated or suspended when the beneficiary state terminates or suspends the agreed reciprocal treatment respectively.⁷⁰ Thus, the right can only be acquired by a beneficiary state upon according the agreed reciprocal treatment to the granting state. This denotes that, unlike GATT I:1, the EU-ECOWAS 16:2 and 3 and EU-EAC 15:1 and 2, MFN treatment under the PTG is not granted automatically, which justifies its omission of the term ‘immediately and unconditionally’ as presented in GATT I:1.⁷¹ Although the EU-ECOWAS and EU-EAC MFN provisions omitted the term ‘immediately and unconditionally’, they do not restrain MFN automaticity when preferential treatment is offered to a third party within the scope of their MFN application. Also, despite the fact that the MFN provisions in both treaties make room for consultation in certain specific instances, the outcome of such negotiation is not subject to a reciprocal exchange of concession.⁷² The non-automatic nature of the PTG MFN clause is further implied by the usage of the terms ‘each other’ in Article 18:1 of the agreement and ‘one another’ in Article 4:1 of the PTG instead of ‘all’.⁷³ These depict that the treatment is to be accorded to only state parties that meet the reciprocal condition and not to ‘all’ state parties on the mere basis of being a party to the treaty.

What this MFN clause promises is to provide an avenue for state parties to benefit from similar treatments accorded third parties states upon agreeing to offer same.⁷⁴ Bailey equates this system to a hotel where ‘all are welcome on payment of the same published charges’.⁷⁵ It aims at establishing material reciprocity by ensuring symmetry between the benefits provided by the granting and beneficiary states. The implication here is that, in certain instances, some state parties may be eligible for MFN treatment while others will not. It is therefore established that the PTG MFN clause, unlike the GATT, EU-ECOWAS, and EU-EAC, is materially reciprocal in nature and hence non-automatic.

4.4 Product Likeness

The MFN provisions under the PTG declined the use of the ‘like’ term and therefore may seem to strip off the need to establish product likeness when granting MFN treatment. The ‘like’ term remains a common feature of MFN clauses and has been instrumental in the outcome of various litigations. In the case of Canada–Autos, the Appellate Body (AB) concluded that Article I:1 of GATT requires that:

any advantage, favour, privilege or immunity granted by any Member to any product originating in or destined for any other country shall be accorded immediately and unconditionally to the like product originating in or destined for the territories of all other Members.⁷⁶

Hence, per the body's assessment, Canada did not accord the MFN treatment immediately and unconditionally to ‘like’ products from or destined for the territories of all other members which constitutes a breach of GATT I:1.⁷⁷

Though the provisions of the rules of origin will determine products that qualify for preferential treatment under the AfCFTA, some may argue that the omission of the ‘like’ term in the PTG MFN provisions would raise difficulty in substantiating claims of discrimination among products and at the same time aggravate a panel's assessment of products' comparability. This may nurture contentions that once preferential treatment has been granted to a product, other state parties have the right to demand same preference for any of their products. However, such contentions will be rendered premature since even in the absence of an explicit ‘like’ term, the general rule is that state parties may raise MFN claims only for products specified in the clause or within the same category as the product being offered the favourable treatment on the basis of which the claim is being made.⁷⁸ By implication, even though the PTG MFN provisions omitted the ‘like’ term, the legal concept of product likeness persists. For simplicity and the avoidance of discrimination, such determination may be made using the intrinsic characteristics of the products concerned rather than external features.⁷⁹ It however remains ambiguous as to why the drafters of the AfCFTA Agreement chose to maintain the ‘like’ term in the National Treatment clause but omitted it from the MFN provisions.⁸⁰ This may raise questions as to whether the omission was intentional on the part of the drafters. However, whether the omission was deliberate or not, the above analysis establishes that it has no impact on MFN application as the ‘like’ term remains implied.

4.5 Third-Party Agreements

With regard to third-party trade agreements predating the AfCFTA, Article 18:3 of the agreement guarantees their continued existence by stressing that the agreement ‘shall not nullify, modify or revoke rights and obligations under preexisting trade agreements that State Parties have with Third Parties’.⁸¹ Thus, they continue to be as effective as they were prior to the AfCFTA as long as they do not ‘impede or frustrate’ its objectives.⁸² Article 18:2 provides additional conditions for this continued existence. A state party is obliged to offer other state parties the ‘opportunity to negotiate’ preferences granted under its third-party agreements that existed prior to the AfCFTA.⁸³ Such preferences are to be offered based on the principle of reciprocity.

The term ‘opportunity to negotiate’ denotes the fact that the MFN clause does not permit an automatic importation of third-party treaty preferences but rather avails a platform for negotiating such preferences on a reciprocal basis. This is affirmed by Article 4:4 of the PTG which provides that a:

State party shall not be obliged to extend to another State Party, trade preferences extended to other State Parties or Third Parties before the entry into force of the Agreement. A State Party shall afford opportunity to the other State Parties to negotiate the preferences granted therein on a reciprocal basis, taking into account levels of development of State Parties.⁸⁴

The above provision exerts control on the use of the MFN clause in the extension of preferences and has the potency of making dispute outcomes more predictable. For this clause to be triggered, a state party must, first of all, indicate an interest in such preferences and must be willing to meet the reciprocal condition. Once the clause is triggered by a state party, the ‘opportunity to negotiate’ must be automatically offered to all other state parties on a reciprocal basis. During such negotiations however, the development levels of state parties must be taken into consideration. This implies that, during such negotiations, obligations taken in exchange for preferences might not be equal as a result of developmental differences among state parties.

The provisions do not preclude an uninterested state party from turning down the ‘opportunity to negotiation’ when offered. This also implies that state parties have the freedom to decide whether to accept or reject the negotiation offer. However, as to whether the preferences negotiated under this clause constitute a new agreement or become part of the AfCFTA Agreement is presently unknown. In addition, as to whether the preferences to be negotiated include the dispute settlement provisions remains unclear. As far as dispute settlement is concerned, the only limitation under the Protocol on Rules and Procedures on the Settlement of Disputes is that: ‘a state party which has invoked the rules and procedures of this protocol with regards to a specific matter, shall not invoke another forum for dispute settlement on the same matter’.⁸⁵ There is also the question of how the ‘levels of development of state parties’ will be accounted for since the agreement left the term undefined and without any mode of assessment.

For new trade agreements established after the inception of the AfCFTA, Article 4:2 of the PTG provides that:

Nothing in this protocol shall prevent a State Party from concluding or maintaining preferential trade arrangements with Third Parties, provided that such trade arrangements do not impede or frustrate the objectives of this protocol, and that any advantage, concession or privilege granted to a Third Party under such arrangements is extended to other State Parties on a reciprocal basis.⁸⁶

Thus, new preferential trade agreements are permitted as long as they do not ‘impede or frustrate’ the objectives of the AfCFTA, and preferences offered by such arrangements are accorded to state parties reciprocally. This appears quite different from the regulation concerning preexisting agreements. As noted above, with regard to third-party agreements in existence before the AfCFTA, the granting state is under no obligation to extend such preferences to other state parties. Instead, it is to afford state parties the ‘opportunity to negotiate’ for the preferences. In the case of agreements entered into after the inception of the AfCFTA, the granting state is under the obligation to extend such preferential treatments to other state parties on a reciprocal basis.

Fitting this into the real-world context, the provisions here can be triggered against Mauritius' recent agreements with China and India.⁸⁷ An AfCFTA state party may request the same treatment it deems more preferential under any of the two agreements mentioned. Upon request for the preference, however, the beneficiary state must consult Mauritius and decide on what it considers ‘equivalent’ to what China or India is offering in return for the preference granted. As noted by the ILC, a state can only enjoy the benefit of an MFN clause conditioned on reciprocity, ‘after assuring the granting State that it will accord to it … treatment of the same kind’.⁸⁸ Therefore, until an equivalent treatment is agreed upon between the parties involved, the clause cannot be applied.

In comparison with the EU-EAC and EU-ECOWAS, the PTG MNF adopts a distinctive approach to third-party agreements. Besides being reciprocally conditioned, it is applicable to both pre and post AfCFTA agreements while the EU-EAC and EU-ECOWAS are only applicable to third-party agreements postdating them.⁸⁹

4.6 Intra-African Relations

Per Article 18:1 of the AfCFTA Agreement, the MFN clause can mainly be triggered in relation to third-party states. That notwithstanding, in recognition of existing and prospective future inter-state party arrangements, provisions were made for such agreements. Although preexisting African Regional Economic Communities (RECs) were not expressly captured under the MFN provisions, Article 19:2 of the agreement guarantees their continued existence. It states that:

State Parties that are members of other regional economic communities, regional trading arrangements and custom unions, which have attained among themselves higher levels of regional integration than under this Agreement, shall maintain such higher levels among themselves.⁹⁰

This is equally captured in Article 8:2 of the PTG which states that:

Notwithstanding the provisions of this Protocol, State Parties that are members of other RECs, which have attained among themselves higher levels of elimination of customs duties and trade barriers than those provided for in this Protocol, shall maintain, and where possible improve upon, those higher levels of trade liberalisation among themselves.⁹¹

The terms ‘shall maintain such higher levels’ and ‘shall maintain, and where possible improve upon, those higher levels’ denote that legal provisions under preexisting regional agreements with a lower integration effect than the AfCFTA are to be superseded by it. This is captured in Article 19:1 of the agreement which stipulates that in an event of inconsistency with any regional agreement, the AfCFTA prevails unless otherwise provided. The term ‘among themselves’ in Article 19:2 of the agreement as stated above may be argued to indicate that regional economic agreements are immune to MFN application under the AfCFTA.⁹² Such a premise may however be unpersuasive as none of the legal provisions expressly bars AfCFTA state parties from seeking preferences granted under the existing regional economic agreements. This suggests that preexisting regional agreements are not immune to MFN application. In fact, Article 4:3 of the PTG encourages higher trade liberalization than what is presently availed by the AfCFTA. Inference from Article 4:4 of the PTG provides a framework for initiating such requests. The granting state in such instances is however under no obligation to extend preferences but rather to offer the beneficiary state an opportunity to negotiate for the preferences on a reciprocal basis.⁹³

Being parties to the AfCFTA does not prohibit state parties from establishing other preferential arrangements among themselves as long as such arrangements are in tune with the objectives of the AfCFTA and the preferences are extended to other state parties on a reciprocal basis.⁹⁴ It should be noted that for new inter-state party agreements postdating the AfCFTA, the required extension is not an ‘opportunity to negotiate’ but rather the actual ‘preferences’, which indicates the obligatory implications Article 4:3 of the PTG possesses. In this provision, the general term ‘preferences’ was used without any specification as long as they conformed with the objectives of the AfCFTA. The fact that the term ‘preferences’ was used in its general form and not ‘preferential trade arrangement’ or ‘trade preferences’ as used in other clauses in the same Article propels the logical conclusion that the term ‘preferences’ connoted a broader interpretation. This points to the quest for greater regional integration as envisaged by the AfCFTA.⁹⁵ For inter-state party agreements predating the existence of the AfCFTA, members are under no obligation to extend ‘preferences’ but rather to afford the other state parties the ‘opportunity to negotiate’ such preferences on a reciprocal basis taking into account their development levels.⁹⁶ It can be inferred here that as with third-party agreements, the PTG MFN provisions also offer distinctive approaches to pre- and post-AfCFTA inter-state party agreements.

4.7 De Facto MFN Implications

Among the recent debates surrounding MFN treatments has been whether the legal provisions embedded in treaties encompass both de jure and de facto discrimination. De facto discrimination can be described as discriminative measures orchestrated to circumvent the basic purpose of Articles of non-discrimination.⁹⁷ The application of this doctrine has been the subject of some panel conclusions. Article 18:1 of the AfCFTA Agreement states that: ‘Following the entry into force of this Agreement, state parties shall, when implementing this Agreement, accord each other, on a reciprocal basis, preferences that are no less favourable than those given to third parties.’⁹⁸

In the context of de facto discrimination with reference to the above text, emphasis should be placed on the term ‘preferences that are no less favourable’. In the case of EC–Bananas III where the AB was tasked to ascertain whether the term: ‘treatment no less favourable’ constitutes both de jure and de facto discrimination within the context of the MFN obligation under the General Agreement on Trade in Service (GATS) II,⁹⁹ it emphatically noted that ‘there is more than one way of writing a de facto non-discrimination provision’.¹⁰⁰ It was upheld that the ordinary meaning of the MFN clause as provided in GATS II does not exclude de facto discrimination.¹⁰¹ Against this background, the present authors contest that irrespective of the differences between the words ‘preferences’ and ‘treatment’, the above terms are substantially similar and therefore warrant similar interpretation. The AB in the above-referred case concluded that the term ‘treatment no less favourable’ should be ‘interpreted to include de facto, as well as de jure, discrimination’.¹⁰² This ruling stifles states from doing indirectly what they cannot do directly.¹⁰³ Thus, discrimination on both de jure and de facto basis is prohibited. To this extent, the present authors argue that the MFN provisions, as provided in the AfCFTA Agreement, encompass both de jure and de facto discrimination.

A practical application of MFN provisions on de facto discrimination was witnessed in the case of the European Economic Community – Imports of Beef from Canada,¹⁰⁴ where the panel was tasked to examine whether the European Economic Community's (EEC) levy-free tariff quota regulations for high-quality grain-fed beef was consistent with GATT Article I. Per the regulations, the levy-free qualification was subject to the attainment of a certificate of authenticity. The primary contention here was the fact that the only agency authorized to issue the certificate was the Food Safety and Quality Service (FSQS) of the United States Department of Agriculture (USDA), an entity empowered to certify meat originating from the United States. The panel concluded that the regulations were inconsistent with the MFN principle in GATT Article I as it restricted EEC market access from origins other than the United States.¹⁰⁵ Thus, in the panel's view, the fact that the FSQS, a USA agency empowered to certify ‘meat originating from the United States’, was the only authorized agency to issue the qualification certificate implied that meat originating from other states was indirectly being denied access to certification in order to benefit from the levy-free tariff quota regulations. Hence, the ECC's levy-free tariff quota regulations were discriminative in favour of the USA.¹⁰⁶

The AB in Canada–Autos affirmed this position in case law by stressing that:

Neither the words ‘de jure’ nor ‘de facto’ appear in Article I:1. Nevertheless, we observe that Article I:1 does not cover only ‘in law’, or de jure, discrimination. As several GATT panel reports confirmed, Article I:1 covers also ‘in fact’, or de facto, discrimination.¹⁰⁷

To this extent, the present authors acknowledge established international legal practice of curtailing both de jure and de facto discrimination and argue that the MFN provisions of the PTG conform with international practice.

The above analysis demonstrates that the PTG MFN clause, unlike that of the GATT, EU-EAC and EU-ECOWAS goes beyond implied reciprocity as a general feature of MFN clauses to actually establish material reciprocity. Hence, the clause is a conditional one which strips it of the general automaticity of the MFN application. Although it seeks to curtail both de jure and de facto discrimination, the clause can only be triggered upon meeting the reciprocal condition. It is also indicative from the comparative analysis made that African states do not have a single approach to drafting MFN clauses. That said, a careful examination of the EU-EAC, EU-ECOWAS, and the PTG MFN clauses depict a preference for consultation by African states. This is deduced from the fact that resort to consultation is prevalent in the MFN provisions of all the three agreements mentioned.¹⁰⁸ It is therefore indicative that African states prefer to resolve MFN related issues through dialogue than to seek legal redress.

5. Implications

It is undeniable that irrespective of the nature of text encapsulated in an MFN clause, it remains deficient in solving all problems of international trade. Even the widely used unconditional MFN clause has, to some extent, been noted for discouraging greater trade liberalization.¹⁰⁹ Its automatic nature derails countries from engaging in wider tariff concessions as doing so may trigger an automatic extension to other states without the obligation for those states to offer anything in return. Viner alludes to a situation where ‘a country which follows the unconditional practice may deliberately refrain from making concessions to any country in order that it may avoid the obligation of extending such concessions to other countries’.¹¹⁰ For instance, if country K after concluding an agreement with country L proceeds to make concessions with countries M, N, and O, it is through unconditional MFN obliged to automatically extend the privileges under the new concessions to country L. Country L in this instance, may in order to refrain from further obligations, deliberately decide not to conclude any other agreement but rather continuously receive preferences accrued from third-party concessions made by K through unconditional MFN application. Such a mechanism would disincentivize K from making further trade concessions since, in practice, countries may be ‘generous but not naive’.¹¹¹

The unconditional MFN clause may also serve as a lock-in tool for strategic trade partnerships.¹¹² Thus, the mere prospect of automatically extending favourable concessions to other treaty partners may discourage countries from granting more favourable treatments to new treaty partners. As summarized by an interviewee in Adriaesen and Postnikov, 2022,¹¹³ with unconditional MFN, once you put an offer on the table, ‘you cannot offer anything better to future trading parties’. Its automaticity decreases governments' ability to tackle MFN related unforeseen circumstances, therefore increasing their reluctance to further trade expansion.¹¹⁴ A practical effect of this is the United Kingdom's (UK) recent difficulties in its trade negotiations with the EU where the automaticity of the unconditional MFN clauses featured in both parties' previous concessions with third parties has constrained their ability to negotiate more favourable treatments.¹¹⁵ The unconditional MFN does not only discourage tariff reduction but also serves as a motivation for tariff increment for bargaining purposes.¹¹⁶ States become inclined towards initially imposing very high tariffs only to reduce them under treaties so as to appear committed to those treaties.

To this extent, it is based on reciprocal MFN clauses that state parties will be enthused to expand tariff concessions and curtail other trade barriers.¹¹⁷ They remain confident and content with the fact that they give nothing without receiving anything in return. Thus, it curtails the free rider syndrome. Within the AfCFTA framework, the relevance of the reciprocal MFN clause is most manifested in guaranteeing the preservation of the acquis, the existing RECs’ Free Trade Agreements (FTAs) which form the foundation upon which the agreement is built.¹¹⁸ The term acquis was first introduced into the African integration process during the Tripartite Free Trade Area (TFTA) negotiations which served as a roadmap to the AfCFTA negotiation.¹¹⁹ Available AU documents establish that the AfCFTA roadmap, was among others, informed by the RECs and Tripartite acquis.¹²⁰ In this regard, although the AfCFTA Agreement does not define its acquis, inference can be made from the TFTA negotiations. During the TFTA deliberations, the implication of the acquis was explained as follows:

Building on the acquis of the existing REC FTAs in terms of consolidating tariff liberalisation in each REC FTA: Acquis is a French term meaning ‘that which has been agreed’. In the context of the Tripartite Free Trade Agreement, it means that the negotiations should start from the point at which the COMESA, EAC and SADC trade negotiations have reached. Tariff negotiations and the exchange of tariff concessions would be among Member/Partner States of the Tripartite FTA that have no preferential arrangements in place between them. This will both preserve the acquis and build on it.¹²¹

Inferring from the above within the context of AfCFTA, the fundamental conclusion is that there would be no AfCFTA without the existing RECs which therefore substantiates the need for their preservation. With this complexity of interlocking existing RECs with a continent-wide FTA, adopting an unconditional MFN would trigger unintended consequences.¹²² It would imply that a member of the Economic Community of West African States (ECOWAS) could through the unconditional MFN claim privileges under the Southern African Development Community (SADC) agreement without the obligation to offer anything in return. Besides the multiple complexities such an agreement would create, it is also likely to scare away national governments knowing the limited control they would have on such future consequences.¹²³ However, the reciprocal MFN, as is currently being applied, affords countries the opportunity to bilaterally negotiate favourable treatments and the assurance that nothing would be given for free.

The conditional MFN clause is asserted to be discriminatory in nature and does not offer any advantage over the unconditional one.¹²⁴ The condition of reciprocity is also noted for raising difficulties in interpretation with regard to identifying which treatment is ‘equivalent’ to that which is being offered.¹²⁵ However, the present authors posit that with the AfCFTA in perspective, such negatives of the reciprocal MFN are unlikely to manifest. For agreements predating the AfCFTA, the term ‘opportunity to negotiate’ as used in the provisions gives the indication that before such preferences are extended, state parties concerned will have the opportunity to decide among themselves which treatment constitutes an ‘equivalent’ treatment. Even with agreements postdating the AfCFTA, the extension of preferences is rendered non-automatic by the reciprocal condition.¹²⁶ Meeting this reciprocal condition on the part of the beneficiary state would require a discussion among the parties concerned and, in that sense, an agreement would be reached on what constitutes an ‘equivalent’ treatment.

MFN under the PTG avails a mechanism for state parties to relate to each other individually, which would enable them meet the requirement of ‘taking into account levels of development of state parties’.¹²⁷ This term, as used in the PTG MFN provisions, decimates the argument that the negotiation framework of offering something in exchange for something may be unfair towards states with very few or low duties.¹²⁸ Those states are mostly less developed comparatively. Therefore, although the term ‘taking into account levels of development of state parties’ may not guarantee that less developed states will not be disadvantaged, it at least provides a legal requirement for such considerations to be made during negotiations. This is further reflected in the ‘flexible and special and differential treatment’ as a governing principle of the agreement¹²⁹ and Article 6 of the PTG which provides that:

In conformity with the objective of the AfCFTA in ensuring comprehensive and mutually beneficial trade in goods, State Parties shall, provide flexibilities to other State Parties at different levels of economic development or that have individual specificities as recognised by other State Parties. These flexibilities shall include, among others, special consideration and an additional transition period in the implementation of this Agreement, on a case by case basis.¹³⁰

The use of ‘shall’ stipulates the obligatory nature of the special and differential consideration.¹³¹ This system seeks to pique uniformity in trade relations and promote the predictability of judicial outcomes.

The above provision, like the WTO special and differential treatment provision, also reflects the principle of variable geometry featured in Article 5 of the AfCFTA Agreement.¹³² The principle of variable geometry is a strategy that allows the negotiation of issues to an agreement of which not all are binding on all parties to the agreement.¹³³ It ensures that obligations differ for parties to a treaty.¹³⁴ This concept is further explained by the East African Court of Justice's reasoning that the principle of variable geometry is ‘intended, and actually allows, those Partner States who cannot implement a particular decision simultaneously or immediately to implement it at a suitable certain future time or simply at a different speed while at the same time allowing those who are able to implement immediately to do so’.¹³⁵ Thus, the principle accommodates the differences in readiness among nations to a treaty. Considering the establishment of a continent-wide FTA in the presence of sharp economic diversity among African states, this paper concurs to the opinion of judge Tanaka of the International Court of Justice that ‘to treat unequal matters differently according to their inequality is not only permitted but required’.¹³⁶

The practicability of such special considerations is already exhibited in the AfCFTA's liberalization process. By the negotiation modalities, AfCFTA state parties have agreed to eliminate 90% of tariff lines on goods within ten years for Least Developed Countries (LDCs) and five years for non-LDCs. An additional 7% of tariff lines on sensitive goods are to be liberalized over ten years for non-LDCs and 13 years for LDCs.¹³⁷ This also implies that upon full implementation of the AfCFTA, the PTG MFN provisions may only become necessary in relation to the remaining 3% excluded tariff lines. Although 3% may be minute in numerical terms, it would be wrong to make such inference without recourse to Africa's tariff framework. A study by the United Nations Conference on Trade and Development (UNCTAD), and the United Nations Economic Community for Africa (UNECA) discovered that Africa's entire trade is concentrated in about 30 tariff lines.¹³⁸ Therefore, as noted by His Excellency Mr. Issoufou Mahamadou, the then President of Niger and leader of the AfCFTA process, even a 1% exclusion constitutes a substantial proportion of Africa's tariff lines.¹³⁹ In this respect, the 3% excluded tariff lines that will be affected by MFN application still represent a significant proportion of Africa's trade and, for that matter, requires optimum consideration.

One crucial question that beacons in this progressive liberalization process is: how would the reciprocal MFN clause play out in this framework where during the early years of implementation, the levels of liberalization among state parties would be unequal?

It should be noted that the principle of variable geometry is expressed as an opt-in agreement among a subset of countries where the benefits are restricted to those in the subset.¹⁴⁰ Within this context, it is generally expected that states will not request MFN application beyond the concessions they have made. It is however worth noting that the PTG MFN provisions neither prohibit nor oblige the grant of such requests. Such requests if made would fall under Article 4:3 of the PTG which encourages state parties to extend to one another preferences that aim at achieving the objectives of the AfCFTA.¹⁴¹ It should also be noted that the non-obligatory nature of the clause implies that the decision to grant the requested preferential treatment rests on the discretion of the granting state. Even so, once the preference is granted to a state party, it becomes obligatory to be extended to other state parties on reciprocal basis.¹⁴²

Indeed, the reciprocal condition embedded in the PTG MFN is likely to create a pathway of rights and obligations that differ across AfCFTA state parties. This is because what may be regarded ‘equivalent’ concession may vary in various instances based on the states involved. Although some may view this as a disadvantage considering the voidness in equality of concessions made, this should rather be seen as a positive effect that conforms with the principle of flexibility and special and differential treatment embedded in the agreement.¹⁴³ In line with Article 6 of the PTG,¹⁴⁴ it helps overcome the developmental imbalances among the various regions of the continent which have for long posed a major challenge to African integration.

It may also be argued that the reciprocal MFN clause would result in a situation where imports from foreign markets will be offered more favourable treatments than those from Africa and therefore undermine the continent's integration process. This argument may gain some grounds if strictly viewed from a Pan-African lens. From a global perspective, however, it is rather a substantial advantage to international trade with the implication that AfCFTA state parties are not locked from offering more favourable concessions in future treaties. This is because the reciprocal MFN clause quenches the fear of unintended consequences attached to the unconditional clause.¹⁴⁵ The reciprocal condition ensures that the clause cannot be automatically triggered in pursuit of a state party's future concessions with third parties. For the PTG MFN to be triggered, the granting state must gain some advantages in return which alleviates its worry about future MFN implications.

Even if assessed from the Pan-African standpoint, it is for this reason that during the AfCFTA negotiations, state parties were urged to abstain from third-party negotiations until the AfCFTA came into force.¹⁴⁶ Per the requirements of Article 17 of the agreement, a state party must notify the other state parties of any international trade commitments it makes post-AfCFTA.¹⁴⁷ Thus, before a new agreement comes into force, the other state parties are given an opportunity to scrutinize it.¹⁴⁸ This would ensure that all post-AfCFTA agreements are void of provisions that may be considered a threat to African integration as noted in Article 4:2 of the PTG. With third-party agreements concluded before the AfCFTA, state parties maintain the right to negotiate for equal treatment upon granting equivalent concessions.¹⁴⁹ Unlike preferences granted under other agreements that require the application of the MFN provisions to import, preferential treatments provided by the AfCFTA are to be automatically enjoyed by state parties without any reciprocal conditionality.

Though MFN clauses have been noted for multi-lateralizing bilateral agreements, MFN provisions under the AfCFTA denote a multilateral framework hinged on bilateral negotiations. In the African sense, the reciprocal MFN clause possesses the likelihood of guaranteeing stability in national production and protecting acquired living standards.¹⁵⁰ This however does not imply that the reciprocal MFN clause is a guaranteed catalyst for trade enhancement. The reciprocal condition has been criticized for wielding the potential to reduce the benefits provided by the MFN clause¹⁵¹ since its cardinal focus is to ensure equal treatment among state parties and not necessarily equal treatment between state parties and third-party states.¹⁵²

The relevant issue in MFN discussions is not the rapidness or automaticity of its application but rather its effectiveness in the harmonization of international trade. With this in mind, it would be inadequate to analyze MFN provisions without taking into account the economic foundation and background of parties involved. A quantitative study on events under the WTO Dispute Settlement Understanding (DSU) unveiled that states with larger economies and larger trade flows are much more able to utilize the DSU.¹⁵³ Out of over 600 disputes raised under the DSU, only three were initiated by African countries.¹⁵⁴ African states have demonstrated a low appetite for inter-state dispute resolution. This is largely because they lack the means to deploy such legal frameworks.¹⁵⁵ Certainly, disinterest in dispute resolution is a threat to any international agreement. Acknowledgement of this fact is the basis for which every international agreement features a dispute settlement mechanism. Under the United States–Mexico–Canada Agreement for instance, cases like the Dairy TRQ Allocation Measures, Automotive Rules of Origin, and Crystalline Silicon Photovoltaic Cells Safeguard Measure were triggered under its dispute settlement mechanism.¹⁵⁶

Until African states are able to develop the required means for deploying international dispute resolution mechanisms, it would be rather innovative for them to initiate legal frameworks that minimize the need for dispute resolution. Upon a careful examination of the AfCFTA Agreement, the present authors assert that this is a major objective its MFN provisions seek to accomplish. It establishes a system with the prospect of curtailing the need for MFN related dispute resolution since the clause cannot be triggered without an engagement between the state parties concerned. This assertion is based on the fact that the AfCFTA PTG MFN clause only functions through a system of consultation. As the USA's former Secretary of State, Hughes puts it, what the reciprocal MFN provides is ‘an opportunity to bargain for’ preferential treatment.¹⁵⁷ It is therefore expected that since the reciprocal MFN cannot be applied unless an understanding is reached among concerned parties, there would be little room for MFN related disputes because prospective causes of future contentions are likely to be resolved in the consultation process even before the clause is applied.

6. Conclusion

From the above discussions, it is certain that MFN clauses play a critical role in international trade and possess legal consequences depending on how they are drafted in the treaty concerned. In comparison with GATT, EU-EAC and EU-ECOWAS, the wording of the AfCFTA MFN provisions is more decentralized with separate provisions for pre- and post-AfCFTA agreements (third-party and inter-state party agreements) to dilute legal contentions. It is highly based on dialogue and compromises between state parties. The provisions accommodate a guided bilateral state engagement which considers developmental gaps rather than a holistic approach.

In the African context, this is critical for a continent-wide trade agreement. The idea of an economically integrated Africa has been under consideration since the early Pan-Africanists' days of the 1960s. The vision was, however, deferred by the unwillingness of some states across the continent. Such hesitance continues to linger as some major African economies, including South Africa and Nigeria, were initially reluctant to join the AfCFTA. In defence of its stance, the information minister of the only non-party African state, Eritrea, in a recent post, labelled the agreement as an unachievable ‘abstract platitude’.¹⁵⁸ Even in its present state, though 54 African countries have signed on to the agreement, many are yet to complete the final stages of ratification and submission of tariff concessions, respectively to permit its full implementation. This has compelled an interim adoption of the AfCFTA Initiative on Guided Trade.¹⁵⁹ The initiative is a pilot implementation of the AfCFTA with eight African states to, among other things, demonstrate the feasibility of the agreement.

Against this backdrop, the pessimism among African states necessitates the use of non-traditional measures such as the reciprocal MFN clause to cling members' interest in remaining parties to the agreement. The reciprocal MFN clause enables the granting state to retain some control over MFN application. Hence, the fear of MFN implications when entering into agreements with third-party states is eroded. This is particularly important since intra-African trade constitutes only 16% of the continent's trade activities.¹⁶⁰ Thus, Africa trades significantly more with external parties and, in that sense, AfCFTA MFN implications will be critical to members' extra-continental engagements, taking into account its membership size. The adoption of an unconditional MFN clause would have been a cause of worry to many African states. Indeed, MFN application has in some instances been destructive to the conclusion of future negotiations with third-party states.¹⁶¹ In this regard, Simo notes that the unconditionality of the MFN clause in the AfCFTA Protocol on Trade in Services threatens the conclusion of more favourable third-party agreements (in relation to trade in service).¹⁶² Annulling this fear implies the erosion of MFN related scepticism among AfCFTA state parties when entering into future trade agreements. To this effect, although the reciprocal condition may pose some challenges to MFN application, it constitutes a necessary compromise for the AfCFTA's existence.