Special Economic Zones in an Era of Multilateralism Decadence and Struggles for Post-Pandemic Economic Recovery: Perspectives from the Global South

Abstract

Recent years have seen a proliferation of Special Economic Zones (SEZs) in developed and developing countries. Developed in Europe in its modern shape, most SEZs are located outside the continent today, notably in the developing world, where SEZs form part of these countries’ export-oriented growth policy tools and overall economic development. At a period of growing unilateralism and the return of the State as an economic actor, this contribution seeks to tackle the rise of SEZ laws in the global south, with a particular focus on Africa. It will scrutinize the reasons for their establishment, the measures chosen to promote them, and the international ramifications in these respective regions and broadly on the global plane, notably at the WTO. With the entry into force of the African Continental Free Trade Area (AfCFTA) Agreement, African countries face challenges of multi-layered SEZ governance, which this contribution intends to address. These challenges also extend to the cross-regional trade agreements these countries conclude, individually and as a bloc. Since SEZs are often assimilated with a category of subsidies and are discriminatory trade measures, this contribution, in essence, investigates the extent to which current trade rules at multilateral and regional levels address these controversial aspects of SEZs.

A. Introduction

In August 2017, a World Trade Organization (WTO) panel issued a report in a dispute opposing Brazil against the European Union and Japan.¹ The case originated from a complaint against Brazilian fiscal and regulatory measures in its industrial programs. In particular, some of the measures granted tax exemptions to export-oriented firms. This dispute attracted attention in the literature for many reasons, chiefly because of the acceptance of the Panel that bridging the digital divide and promoting social inclusion constituted public moral concerns capable of justification under Article XX(a) of the General Agreement on Tariffs and Trade (GATT).² Less studied were those aspects of the measures that led to the dispute concerning the tax incentives for goods produced in the Brazilian Manaus Free Trade Zone (FTZ),³ which the Panel failed to entertain consistently with other prior missed opportunities.⁴ However, Special economic zones (SEZs) are relevant and have implications for the multilateral trade regime, even though some panels seem to have shied away from conducting systematic analyses of their WTO compliance.⁵

SEZs, such as the Brazilian Manaus FTZ, have been around for hundreds of years. These are geographically delimited areas within a country’s national borders where the government provides companies with a more favorable regulatory and fiscal regime than the national territory.⁶ Put differently, companies in these SEZs are free from restrictions and regulations otherwise imposed on goods and services moving in and out of that country’s territory. As governments are persistently under domestic pressures to respond to demographic, labor and employment, and broader economic challenges to which they are confronted, seeking and retaining foreign capital has become a vital necessity. Therefore, governments provide SEZs incentives to attract foreign direct investments (FDIs), boost exports, increase the trade balance and alleviate unemployment.

While Shannon SEZ, in Ireland, was the first so-called “modern” zone,⁷ contemporary forms of SEZs were essentially developed in Asia and championed by China.⁸ This move proved relatively successful for China since its phenomenal economic growth is attributed partly to its SEZs. Its exports from SEZs today represent between 20% and 25% of its GDP. Other Asian countries soon followed suit. Although a latecomer, Africa, has recently also witnessed a blossoming of SEZs in a relentless bid to emulate or replicate the export-led growth experienced in Asia.⁹ While many of these SEZs emanate from domestic initiatives when framing their industrial policies, some are clearly China-inspired, if not China-supported, SEZs.¹⁰ Having industrialized themselves by attracting Western firms to SEZs established on their territories, China and other Asian countries today offer their expertise to African countries with the belief that the relocation of certain Asian firms’ activities in Africa is promising for the continent’s industrial development. Supporting the creation of SEZs in Africa undoubtedly form part of the Middle Kingdom’s geopolitical deployment on the continent.¹¹ This combination of economic motives and statecraft conducted an explosion of SEZs in Africa over the past two decades. According to UNCTAD estimates, out of 5383 SEZs in the world in 2019, around 237 were located in 33 of the 55 African Union Member States.¹² Consequently, the popularity of SEZs as an industrial policy tool is prevalent in developed and developing countries.

The Covid-19 outbreak precipitated an exceptional economic shock in advanced economies. Its responses, including the “whatever it takes” approach espoused by states as the new religion in town, induced an unprecedented resort to WTO-inconsistent measures.¹³ The 2022 Russian “special military operation” in Ukraine, which many assimilate to a full-blown armed conflict, and the ensuing inflation (related and unrelated), which threatens Covid-19 economic recovery, also resulted in a series of dubious WTO practices in response.¹⁴ Expected to spearhead post-Covid-19 economic recovery,¹⁵ especially in developing countries, SEZs pose several challenges to international economic governance as an instrument of national industrial policies. Through fiscal and regulatory measures, SEZs are vehicles for countries’ artificially-created comparative advantages.¹⁶ While there is no specific global legal and regulatory framework for SEZs, these incentives are essentially unilateral measures by the establishing countries that may sometimes thwart their international obligations. In particular, while the economic effects of SEZs are acknowledged by some or criticized by others,¹⁷ financial incentives may constitute (indirect) export subsidies, which would go against the rules of the WTO¹⁸ and regional trade agreements (RTAs) such as the AfCFTA Agreement. For RTAs,¹⁹ the issues extend beyond subsidies and also concern the treatment of goods manufactured and services produced in SEZs for preferential treatment purposes. This problem is accentuated by the fact that some SEZs, notably the free (trade) zones, are equipped with transshipment and storage facilities that could efficiently serve as vehicles to dodge a regional trade agreement’s rules of origin.²⁰

At a period of growing unilateralism and the return of the State as an economic actor that further sees industrial policies back in fashion,²¹ this contribution seeks to tackle the rise of SEZ laws in the global south, with a particular focus on Africa. Broadly recognized in the literature as “one of the most notable institutions of unilateral economic law,”²² SEZs underscore the rationale for policy intervention in the market to address its failures. Indeed, as posited by economists, “the market may not lead to either a good allocation of resources among sectors or the appropriate choice of techniques,” so much so that “industrial policies […] are one of the instruments for addressing these market failures.”²³ SEZ measures increasingly take the form of export subsidies, which, as a means of economic interventionism, international economic law tries to regulate.²⁴ This paper will scrutinize the reasons for SEZs’ establishment, the measures chosen to promote them, and the international ramifications in Africa and broadly on the global plane, notably at the WTO. With the Agreement establishing the African Continental Free Trade Area (AfCFTA) having entered into force,²⁵ African countries face challenges of multi-layered SEZ governance, which this contribution intends to address. Since SEZ schemes are often assimilated with a category of subsidies and are discriminatory trade measures, this paper, in essence, investigates the extent to which current trade rules at multilateral and regional levels address these controversial aspects of SEZs.

This paper is structured as follows. Following this introduction, Section 2 introduces the concept of SEZs, their definition and the measures chosen by developing countries to establish them. Section 3 deals with international trade law implications of SEZs, notably with regard to rules on unfair trade and the regime governing goods produced in these zones. Section 4 concludes the analysis.

B. The Concept of SEZs in Global South – Perspectives from Africa

I. Definition of SEZs

SEZs is a generic term with no specific definition in literature or by countries that establish them. African countries use SEZs as an instrument of economic policy to promote FDI by offering a more competitive business environment through the granting of tax, regulatory and financial advantages and a range of services often tailored to the needs of the companies. This is why SEZs take on different models and names depending on their missions, and the objectives pursued. It is generally through a government agency in charge of stimulating investment that governments promote the SEZ incentive package.²⁶ The term SEZ covers several different types, which governments adapt according to their respective development strategies, explaining the wide variety of situations observed worldwide. Moreover, it is widespread for many countries to set up different forms of SEZs simultaneously in order to optimize their positive effects.

The term covers a broad array of zones ranging from free zones (FZs),²⁷ export processing zones (EPZs),²⁸ foreign or free trade zones (FTZs),²⁹ enterprise zones to industrial parks³⁰ (or industrial zones), and others.³¹ The International Labor Organization (ILO) defines EPZs as “industrial zones with special incentives set up to attract foreign investors, in which imported materials undergo some degree of processing before being exported again”.³² For merchandise trade purposes, the central character of SEZs is that “any goods” introduced in that part of the territory (established as an SEZ) “are generally regarded, insofar as import duties and taxes are concerned, as being outside the Customs territory” of that country.³³

Despite their disparate appellations, SEZs share four main standard features.³⁴ SEZs are generally geographically delimited areas, usually physically secured (fenced-in). In Burundi, for instance, the SEZ is located in Warubondo,³⁵ a locality situated in the capital province, Bujumbura, with a dedicated surface area of more than five square kilometers. Another common feature of SEZs is their single management/administration structure. Moreover, eligibility for benefits is generally based on physical location within the zone. Incentives only extend to firms established in the SEZ which are engaged in designated activities, which are usually tied to the objectives behind the establishment of the zone. Lastly, SEZs are separate customs areas that benefit from streamlined procedures.

Therefore, the SEZ concept refers to a plurality of situations whose common point lies in the identical advantages they provide: tax relief for economic activities, relaxed and favorable regulations, and exemption from customs duties.

II. Purposes and Objectives of African SEZs

Since the late 1970s, countries have used SEZs as tools for opening up their markets to international trade and foreign direct investments. In particular, SEZs were used in centrally planned economies to test and accelerate the adoption of market-economy policies. Post-independence Africa also embraced what they conceived then as a powerful lever for their development. Grown in considerable number since the early 1980s, SEZs are notable indicators of the globalization of markets, so much so that attention has shifted from tallying countries that have legislated in that respect to instead considering which countries are still lagging behind.

In 1961, Morocco was the first African country to enact an export processing zone legislation.³⁶ Mauritius,³⁷ Senegal,³⁸ Ghana, and Liberia immediately followed this path. However, it was during the 1990s that the trend spread considerably across the continent. SEZs were initially conceived for developing countries as a “second-best type solution for a country to profit from a greater and more efficient integration into the international division of Labor [sic] without subjecting the whole economy to trade Liberalization and deregulation.”³⁹ This is, however, no longer the case as SEZs now form part of countries’ industrial policies even when they fully commit to World Trade Organization (WTO) rules. In other words, they are no mere alternatives anymore,⁴⁰ as governments prevalently combine trade and industrial policy instruments in their economic growth strategies.⁴¹

The rise of SEZs globally, particularly in the developing world, is primarily attributed to the direct benefits they offer to host countries. These benefits include attracting FDI, job creation, income generation, growth, export diversification and increased foreign exchange earnings. SEZs are supposed to benefit the host country and the investors established there. For the government, SEZs create new job opportunities, permit the development of export-oriented industries, increase export, thus foreign exchange receipts, stimulates foreign investments, tend to contribute to infrastructure development, and can also constitute potential markets for domestic raw materials and natural resources.

Yet, the African SEZ’s objectives are manifold. Not only do they include the economic purposes mentioned above but also broader social concerns for some of them. In South Africa, for instance, besides boosting economic activities, the SEZ law explicitly mentions that it aims at creating “decent work and other economic and social benefits in the region in which it is located.”⁴² As known, the question of decent work forms part of the Sustainable Development Goals (SDGs) to be achieved by 2030,⁴³ which South Africa integrates into its SEZ program. As further elaborated by the SDGs agenda, decent work comes with “full and productive employment and decent work for all women and men, including for young people and persons with disabilities, and equal pay for work of equal value.”⁴⁴ Other “social benefits” of South African SEZ legislation include environmental protection, which is why, as part of incentive packages, an SEZ developer can access SEZ funds for environmental impact improvement post-establishment in the SEZ.⁴⁵

In Rwanda, one rationale for establishing the SEZ is the promotion of “a high-quality business climate with an emphasis on environmental protection.”⁴⁶ To this end, Article 39 of Rwandan SEZ legislation states that developers (i.e. individuals or legal entities with a license to establish and develop an SEZ), operators (i.e. entities licensed to operate an SEZ), and users (licensees to carry out activities in an SEZ) must comply with laws determining modalities for protection, conservation and promotion of the environment.⁴⁷ In the Democratic Republic of Congo (DRC), prior environmental and social impact assessments, including mitigation plans for these impacts, precede the designation of a special economic zone.⁴⁸ The same holds for Kenya, where a proposed project’s “environmental and social impact” determines the issue of a license to operate in the SEZ.⁴⁹ Consequently, sustainable development is an essential value in African SEZ schemes.

III. Measures Chosen to Promote SEZs in Africa

While the most frequently cited benefits of SEZs are the increase in foreign exchange earnings through export and FDI, SEZs in Africa, to many commentators, are, except for a few, nothing more than white elephants.⁵⁰ Despite the negative view surrounding African SEZs, it is common for governments to provide established companies in SEZs with corporate and business tax exemptions for several years. Likewise, SEZs also have in common the exemption from duties they grant to goods coming from outside the customs territory of the implementing government for as long as they remain in the SEZs before an eventual re-exportation.

Fiscal incentives usually encompass tax breaks, tariff reductions or duties exemptions, and in some instances, exemptions from export taxes. It is worth noting that it is widespread for governments to combine these incentives to ensure a robust result of the policy objectives of the SEZ schemes. Tax breaks are the measures par excellence of SEZ incentives. They come in the forms of profit, corporate, income and sales tax relief, as well as repatriation of profits. As will be discussed below, tax relief constitutes revenue foregone which enters into consideration in the characterization of a measure as a subsidy. From a trade point of view, a tax break is considered to be “essentially a gift from the government, or a waiver of obligations due, and it is clear that the market does not give such gifts.”⁵¹ This is because tax relief makes the recipient better off than it would have otherwise been absent the measure since it would have been subject to the country’s regular taxation regime.

Tariff reductions or exemptions usually take the forms of soft duty drawbacks/exemptions for imports/ VAT refunds for imports, and even export duties exemptions. A duty drawback is a refund of import duties paid on inputs when the final product is exported. In a way, they help producers located in SEZs to be competitive on the global market since the drawbacks, in effect, reduce production costs. The Gambia, for instance, combines many of these incentives. They include exemption from import and excise duties on goods produced within or imported within the SEZ and also exemption from corporate tax and municipal tax.⁵² These incentives, some of which would also qualify in themselves as subsidies, are contingent upon the beneficiaries’ exportation of “at least” 80% of their outputs.⁵³

Non-fiscal incentives, on the other hand, are of many forms, including direct transfer of funds, the reimbursement of transport costs for exports, etc. Of these, direct payments are prima facie very controversial. Direct payments, whether in the form of grants or otherwise, have been found to confer a benefit to the recipient, similar to revenue foregone, thus potentially amounting to a subsidy. The reason is that direct money transfers place the recipient in “a better position” than it otherwise would have been in the marketplace.⁵⁴ While instances of direct payments are rare, Tunisia is one of the African countries providing direct pre-establishment support to SEZ firms in the form of a “premium” when a firm engages in anti-pollution endeavors in relation to its activities within the SEZ.⁵⁵

One also notes the provision of infrastructure and other services at below cost, such as infrastructure development, warehousing facilities and preferential land rental. In Nigeria, for instance, the SEZ law provides for rent-free land at the construction stage, with the normal rent to be determined post-establishment.⁵⁶ Other non-fiscal incentives stem from the provision of a special regime for labor relations, simplified commercial procedures related to imports (the manifestation of which includes no import or export licensing required and no quantitative restrictions), to simplified procedures to set up commercial activity.

In summary, SEZs generally come with three types of incentives to companies, namely services (transport, telecommunication) of a higher quality than what obtains in other parts of the countries, waiver of import and export duties between goods coming and exiting SEZs, and exemption of profits from corporate and income tax.

C. International Trade Law Implications of Global South SEZs

SEZs as instruments of industrial policies have implications on the establishing countries’ international obligations even though SEZs have been around for centuries. International rules can either facilitate these measures or constrain them by requiring that they should be abolished. While international investment agreements (IIAs) usually act as facilitators,⁵⁷ for instance, by providing favorable conditions for investors, support measures, often to attract FDI under an IIA, may not always sit well with WTO rules. Although no WTO-covered agreement explicitly addresses SEZs, incentives granted under these schemes may fall under WTO rules, notably on subsidies. Measures taken to promote SEZs, including goods produced in them, equally fall under regional trade agreements provision. The consistency of developing countries’ SEZ programs with global and regional trade rules is, thus, worth analyzing.

I. Trade in Goods Manufactured in SEZs for Preferential Tariff Treatment

Besides the question of subsidies, analyzed later, as an industrial policy tool, SEZs also require attention concerning goods not necessarily destined for export outside the regional trade agreements to which the implementing country belongs but destined for the regional market. While the issue is relatively settled under domestic law, since goods produced under SEZs are generally destined for export and may only enter the customs territory of the establishing country after paying the usual import duties,⁵⁸ the question is fairly different when the same goods can be sent in the market of partner states under an RTA.

Indeed, through SEZ incentives, notably tariff exemptions, foreign firms may be inclined to access the RTA domestic markets through tariff-jumping. In other words, SEZ-produced goods could be exempt from import duties as though they are obtained in the territory of one of the state parties to the RTA, absent appropriate measures. The reduction of tariffs among the RTA partners translates into the reduction of export costs to firms located in these countries, a situation that may also benefit foreign firms established in an RTA member’s SEZ, which are already benefitting from production costs in the form of incentives. Conversely, disallowing goods obtained in an SEZ from the RTA tariff exemption may also act as a disincentive for the localization of extra-RTA firms in an RTA country’s SEZ, thus defeating the very purpose of their creation. While a too-generous rule may place local producers at a disadvantage vis-à-vis SEZ-established firms, a too-stringent regulation also comes with drawbacks.

1. Rules of Origin and their Importance for Goods Produced in SEZs

The treatment of goods manufactured in SEZs for preferential tariff treatment’s purpose is one of the challenging issues confronting RTAs. From the earlier discussion, the question of whether these products benefitting from incentives can still compete “fairly” with other goods in the free trade area is legitimate. The central issue is the “origin” of goods that can benefit from preferential tariff treatments and whether goods produced in SEZs should be excluded from tariff preferences and thus considered as any other third-country product.

However, there is no one-size-fits-all approach for treating goods manufactured in SEZs in trade agreements. According to the World Customs Organization’s (WCO) guidance on free trade zones, “some” FTAs, on the one hand, include a special provision which excludes products obtained in the SEZs from gaining the origin status to be eligible for preferential tariff treatment.⁵⁹ The exclusion of SEZ-manufactured goods is guided by the arguments posited earlier, consisting of considering these goods as somewhat “subsidized”, for they benefit from fiscal and other incentives that reduce the cost of production. Another reason to maintain a restrictive approach to Rules of Origin (RoOs), including for SEZ-manufactured goods, is to limit the opportunity for freeriding. The United States has been particularly erratic in this area. For instance, before the North American Free Trade Agreement (NAFTA) renegotiation and the passing of the United States-Mexico-Canada Agreement (USMCA) Implementation Act in the US, goods manufactured in the US FTZ did not confer origin as US-manufactured goods for preferential tariff purposes.⁶⁰ Indeed, the US NAFTA Implementation Act provided that goods produced in US FTZs were not to be treated as originating in any NAFTA country upon entry into the US customs territory.⁶¹ The new USMCA Implementation Act,⁶² which repealed the former Act,⁶³ initially scrapped this restriction before making a U-turn only eleven months later. Indeed, under the January USMCA Implementation Act 2020, FTZs-originating goods were no longer denied preferential tariff treatment.⁶⁴ The reason for this paradigm shift seemed at variance with the Trump Administration’s views and desire to “stop the bleeding” from factory closures, job losses and trade deficits.⁶⁵ One would have expected somewhat stricter RoOs for SEZ-produced goods to keep in the spirit of NAFTA’s renegotiations objectives.⁶⁶ From the turn of things, however, the removal of this relaxation resulted from a drafting mistake immediately addressed by making a set of “technical corrections”.⁶⁷ Therefore, a few months later, this restriction disqualifying SEZ-manufactured from UMSCA’s preferential tariff was restored,⁶⁸ thus reflecting the US’s enduring position under NAFTA.⁶⁹

On the other hand, “many” FTAs in the world grant originating status to products manufactured in SEZs and have an explicit provision for the inclusion of SEZs in the FTA.⁷⁰ FTAs do so because the general understanding is that goods produced in SEZs are eligible for tariff preferences as they are originating goods in the territory of the FTA’s contracting parties and meet the applicable origin criteria.⁷¹ Nevertheless, the guidance adds that even in this case of admission of SEZ-made products as originating, “the territorial definition” of an SEZ “may impact on eligibility to benefit from a preferential tariff treatment”.⁷² Since FTAs generally apply to the customs territory of its parties, an explicit FTA provision is required to include goods produced in SEZs in cases where national legislation defines an SEZ as being outside its customs territory.⁷³ How domestic customs laws define SEZs will impact the treatment of goods produced in them, especially because the Revised Kyoto Convention considers that “any goods introduced [in SEZs] are generally regarded, insofar as import duties and taxes are concerned, as being outside the Customs territory.”⁷⁴ In other words, the Revised Kyoto Convention stipulates that goods located in SEZs are considered outside a country’s Customs territory concerning import duties and taxes. The corollary of this situation is that goods sold by a company established on the domestic market to an entity established in an SEZ are considered an export for duties and tax purposes. States could, therefore, establish SEZs to produce goods that may be denied origin status, thus excluded from the tariff liberalization schemes in the FTA.

Rules of origin are, therefore, of utmost importance in this scenario, and the tendency seems to favor granting originating status to goods manufactured in SEZs even though some countries’ practices, such as the US, are far from the epitome of consistency. Rules of origin refer to the specific rules applied by a country or group of countries for assigning national origin to a product. The WTO Agreement on Rules of Origin refers to them as “laws, regulations and administrative determinations of general application applied by any Member to determine the country of origin of goods.”⁷⁵ An attributed origin may designate a country, even a region or part of a country, depending on the circumstances. Origin can also designate a group of countries, for example, within the framework of a customs union or a free trade area. The rules of origin thus make it possible to apply to each product a tariff, any restrictions, specific trade defense rules (such as anti-dumping or countervailing duties), and, where applicable, tariff preferences or exemptions agreed between the trading partners. Conversely, they prevent non-qualifying originating goods from benefiting from the preferential tariff treatment offered under a trade regime.

It follows that the origin of goods is essential for countries’ customs purposes and the functioning of a regional trade agreement. Countries’ tariff schedules usually set different rates for the same products depending on their origin. While these differentiated rates would typically violate the cardinal MFN principle, they could result from preferential tariffs accorded to developing countries to implement the Generalized System of Preferences (GSP) scheme⁷⁶ or a WTO waiver,⁷⁷ constituting a somewhat “positive” discrimination. Rules of origin further find their importance in the implementation of trade policies, notably, as relevant for this paper, in the field of trade remedies.⁷⁸ RoOs will therefore serve as the basis for the correct determination of the countervailing duties to be imposed on goods originating in SEZs if they are found to contravene the subsidies rules.

For regional trade agreements and goods produced in partner countries’ SEZs, RoOs define the conditions to which the goods traded between the state parties must comply before they can benefit from the exemption from customs duties on imports. They are referred to as “preferential” RoOs as opposed to “non-preferential” RoOs in the context of the WTO. Preferential RoOs are particularly important in Free Trade Areas (FTAs) and less so in Customs Unions with a common external tariff regardless of the point of a good’s entry.⁷⁹ In FTAs, preferences are only accorded to goods originating in a state party to the FTA.⁸⁰

International trade is predicated on granting preferential tariffs, whether on a global MFN basis, unilaterally in the context of the GSP, or bilaterally in a reciprocal trade agreement. RoOs ensure that advantages granted to goods originating in beneficiary countries do not benefit imports from third countries. Third countries could be tempted to fraudulently confer on goods produced in their territory, the origin of a state benefitting from tariff preferences. Likewise, RTAs might want to exclude goods manufactured in SEZs from preferential tariff treatment for many reasons, including the protection of their own domestic industries and the preservation of fair competition in the market of a regional trade agreement.

2. African RTAs and the Treatment of Goods Produced in SEZs

African countries face the challenge of SEZs regulation at national, regional and multilateral levels. While industrial policies like this usually garner domestic approvals, for they address economic and social challenges, participation in RTAs could appear as an obstacle. This is because, depending on the framing of the rules, goods originating in one country’s SEZ may be denied preferential tariff treatment in an RTA, thus considered as goods from a third state subject to the default/global MFN tariff. The support for such exclusion may result from the sentiments among African policymakers that established firms in African SEZs are far from being genuinely African.⁸¹ Since most of the capital is foreign, the reasoning continues, they do not always benefit the local economies and should equally not benefit from preferential tariff liberalization. However, denying such a benefit would also renege on one of the motives of creating what is referred to in the African trade agreement context as “regional economic communities” (RECs). Likewise, not all firms established in SEZs are foreign-owned. Denying originating status to all SEZ-produced goods would, therefore, also penalize national and other African investors.

The formation of African RECs responds to political and economic imperatives. Understanding these rationales is essential for grasping why the fate of goods produced in SEZs for preferential tariff treatment matters. While from a political standpoint, a united Africa free from colonial bondage was the prime motive, achieving a level of industrialization capable of integrating African economies into the global trading system underpins the formation of RECs.⁸² Of the many RECs created in the aftermath of independence,⁸³ only eight are recognized by the African Union as building blocs toward the concretization of the African Economic Community.⁸⁴ They are the Arab Maghreb Union (AMU),⁸⁵ the Common Market for Eastern and Southern Africa (COMESA),⁸⁶ the Community of Sahel-Saharan States (CEN-SAD),⁸⁷ the East African Community (EAC),⁸⁸ the Economic Community of Central African States (ECCAS),⁸⁹ the Economic Community of West African States (ECOWAS),⁹⁰ the Intergovernmental Authority on Development (IGAD),⁹¹ and the Southern Africa Development Community (SADC).⁹² Not all of these RECs contain provisions relating to goods produced on SEZs. Practice is inconsistent, even though the tendency seems to favor granting origin to SEZ-produced goods.

Arab Maghreb Union

The AMU Treaty provides for the gradual removal of obstacles to trade in goods.⁹³ In light of this aim, Arab Maghreb Union countries concluded a Trade Protocol in 1991 where they agreed to remove customs duties, taxes and charges having equivalent effects imposed on imports of products originating in their respective jurisdictions.⁹⁴ While containing rules of origin, this Protocol does not address SEZ products but goods wholly obtained and those that undergo a substantial transformation.⁹⁵ Therefore, the AMU Trade Protocol is silent on the treatment of SEZ-made goods.

COMESA

The COMESA Treaty requires goods originating in the Member States to be eligible for common market treatment. It refers to the Protocol on Rules of Origin for the definition of these products.⁹⁶ Like any RTAs, the COMESA Protocol on Rules of Origin sets the criteria for distinguishing goods produced within the Member States and the others for preferential tariff treatment. Concerning goods produced in COMESA Member States’ SEZs, Article 2.12 of Part II-Chapter 2 of the COMESA RoOs Protocol stipulates that they “shall be granted preferential tariff treatment if they meet the requirements of the COMESA Rules of Origin.”⁹⁷ COMESA RoOs provide the classic criteria that goods must meet to qualify as originating, namely wholly obtained and substantial transformation.⁹⁸ If goods meet these requirements, they will be treated as originating and will benefit from preferential tariff treatment regardless of their being made in an SEZ.

East African Community

The EAC Treaty provides for a progressive establishment of an East African Customs Union and Common Market and invites the Partner States to eliminate, among others, tariff and non-tariff barriers on goods.⁹⁹ The Protocol on the establishment of the East African Customs Union was adopted to give meaning to this aspiration. Concerning goods that will benefit from the removal of trade barriers between Partner States,¹⁰⁰ the Protocol stipulates that only originating products that meet the RoOs provisions qualify.¹⁰¹ This is what the EAC Customs Union RoOs were devised to clarify.¹⁰² In attributing originating status to wholly produced¹⁰³ and substantially transformed goods,¹⁰⁴ the EAC RoOs do not treat goods produced in SEZs differently from other products originating in Partner States. This is a change of paradigm from the 2005 EAC Customs Union Regulations, Annex VII on Export Processing Zones, which denied originating status to SEZ-made goods and treated them as any other imports in the EAC customs union.¹⁰⁵

Interestingly, however, EAC RoOs do not confer originating status to final goods whose inputs from other Partner States were subject to subsidies regardless of the amount of subsequent work and processing.¹⁰⁶ This exception may apply in an SEZ context if the goods manufactured in them are later used as inputs for goods produced outside SEZs. The latter will not benefit from preferential treatment on the final goods if SEZ-made inputs meet the EAC subsidy threshold. In other words, subsidized inputs in an SEZ deprive outputs of preferential treatment under the EAC customs union.

IGAD

Regarding the IGAD, its objectives include the harmonization of its Member States’ trade policies, including to “promote and realize the objectives of the [COMESA] and the [EAC].” IGAD’s membership is indeed made of some COMESA states, which are, at the same time, EAC states, so much that one can seriously question, at least from a trade liberalization point of view, the raison-d’etre of the IGAD as a separate entity.¹⁰⁷ As of the date of this writing, IGAD does not have a robust trade instrument for harmonizing trade and customs policies. It follows that most of the goods traded by its Member States among themselves are done under the COMESA or the EAC regime, including how these regimes deal with SEZ-produced goods.

ECOWAS

ECOWAS is the only African RTA among the AU-recognized RECs¹⁰⁸ explicitly refusing to grant preferential tariff treatment to goods originating in SEZs.¹⁰⁹ Article 7, dealing with “goods produced in free zones or under special economic regimes”, states precisely the following:

“Goods transformed within the framework of economic or suspensive Customs regimes or certain special regimes involving the suspension or partial or total exemption from Customs duties on inputs shall in no case be considered as originating products.”

It follows that, regardless of the local content of the final product, it will not be recognized as originating when obtained in an SEZ.

SADC

Under SADC RoOs, a product is considered as originating in a Member State if it meets one of the following criteria: wholly obtained, substantial transformation, and change in tariff heading of the output from non-originating input.¹¹⁰ Like previous RECs, SADC RoOs do not distinguish SEZ-made products from the rest for preferential tariff purposes. One can therefore conclude that goods made in SADC Member States’ SEZs are treated the same as non-SEZ-produced goods.

3. RoOs for Goods Produced in SEZs under the AfCFTA

With the creation of the AfCFTA, the question of rules of origin and goods manufactured in SEZs acquired another level of complexity for African states. AfCFTA’s main objective is the creation of a single continental market for goods and services, which, associated with the free movement of business persons and investments, will pave the way for establishing a continental Customs Union.¹¹¹ With a market spanning 54 African Union Member States, the AfCFTA is today the largest free trade area in the world in terms of membership, with the potential to also become one of the largest integrated markets in terms of volume of trade when fully implemented. The progressive elimination of tariffs and non-tariff barriers to trade in goods is one of the means to achieve a united continental market.¹¹² Adopting an AfCFTA Agreement’s Protocol on Trade in Goods serves to achieve a liberalized market for trade in goods.¹¹³

As an FTA, AfCFTA Agreement also provides RoOs in goods and services. The services RoOs govern the conditions to benefit from service liberalization as a service supplier established in one State Party, whether by being a national or a permanent resident. ¹¹⁴ For the AfCFTA Agreement, goods will be eligible for preferential treatment only if “they are originating in any of the State Parties” in accordance with RoOs conditions and criteria and other product-specific rules to be developed.¹¹⁵ These “criteria and conditions” stipulated in Article 13 of the Protocol on Trade in Goods are located in Annex 2 of the Protocol on Trade in Goods.¹¹⁶ This is also where the definition of an SEZ is provided. Indeed, Article 1(u) of Annex 2 to the AfCFTA Protocol on Trade in Goods describes them as “Special Economic Arrangements” or “Special Economic Zones”. It then defines them as “special regulatory provisions applicable in a geographical demarcation within a State Party’s Territory where the legal, regulatory and fiscal and Customs schemes, applicable to business differ, generally in a more liberal way, from those in application in the rest of that State Party’s Territory”.¹¹⁷

For goods produced in SEZs, the AfCFTA Agreement’s Protocol on Trade in Goods addresses them as a “complementary” policy in Part VII of the Agreement (including “infant industries” policies¹¹⁸ and “state trading enterprises” policies).¹¹⁹ The Agreement recognizes the right of AfCFTA State Parties to “support the establishment and operation of special economic arrangements or zones for the purpose of accelerating development.”¹²⁰ The developmental aspect of SEZs is therefore acknowledged. However, for goods in SEZs, the Protocol on Trade in Goods seems to distinguish, although elusively, between products “benefitting from” SEZs¹²¹ and the trade of products “manufactured” in SEZs.¹²² For the former, i.e. goods “benefitting” from SEZs, the regulations will intend to address how they should be treated, while the latter deals with how they would be traded.

For products “benefitting” from SEZs, Article 23(2) of the AfCFTA Agreement’s Protocol on Trade in Goods states that the Council of Ministers will develop appropriate regulations, which “shall be in support of the continental industrialization programmes [sic]”. Therefore, SEZs form part of the continental industrial programs and the treatment of goods in them is crucial to realizing these objectives. The development of these regulations constitutes outstanding issues under the RoOs negotiations.¹²³ For goods “manufactured” in SEZs, their trade within the AfCFTA is subject to RoOs.¹²⁴ In this regard, Article 9(1) of Annex 2 of the Protocol on Trade in Goods stipulates that they “shall be treated as originating” goods if they meet the requirements of the RoOs. Moreover, State Parties are required to “take all necessary measures” to ensure that goods which are traded under cover of proof of origin and which, during their transportation use an SEZ situated in their territory, “shall remain under the control of the Customs Authority and are not substituted by other goods”.¹²⁵ In other words, Customs Authorities shall ensure that goods that transit through SEZs are not substituted with goods manufactured in SEZs or brought from other places simply for the purpose of being shipped to the destination market. Goods smuggling is known to be a real problem within SEZs,¹²⁶ hence the imperative of strict and effective customs controls. However, transiting goods in SEZs may require handling necessary to preserve the commodities in good condition, which would be tolerated.¹²⁷

Interestingly, an outstanding provision addresses a situation where a product is imported from a State Party into an SEZ. This is generally the case for inputs for the manufacturing of other products. The end product would eventually qualify as originating in the SEZ subject to undergoing substantial processing or transformation per the RoOs criteria.¹²⁸ This would be in contradiction with EAC RoOs that deny originating status to final goods whose inputs come from an SEZ.¹²⁹ Criteria for determining the origins of goods under the AfCFTA RoOs are classic ones. The first is the “wholly obtained” product in a State Party,¹³⁰ and the second criterion is the “substantial transformation” in a State Party.¹³¹ It is out of the scope of this paper to study these criteria in detail. Suffices it to say that goods manufactured in SEZs must meet these criteria to be treated as originating in one State Party and benefit from preferential tariff treatment. The AfCFTA Agreement, therefore, treats SEZ-produced goods as originating, as most African RECs studied earlier.

As mentioned above, African countries face challenges in treating SEZ-produced goods not only in the AfCFTA (continental) but also in the RECs (regional). The AfCFTA’s regime thus adds another layer of complexity to the matters. FTAs created by RECs are recognized as AfCFTA’s “building blocs”,¹³² thus a potential for overlap and even conflict of RoOs and treatment of SEZ-produced goods if not carefully framed and implemented. There is no apparent conflict between AfCFTA RoOs and many African RECs FTAs RoOs concerning the treatment of goods produced in SEZs. They agree that goods produced in SEZs will be granted preferential tariff treatment as any goods manufactured in the “territory” of State Parties. One can, therefore, not anticipate significant legal problems with the free movement of goods produced in African countries’ SEZs under AfCFTA rules.

One such challenge lies with ECOWAS RoOs that do not recognize SEZ-produced goods as originating products and the EAC RoOs that exclude outputs made from subsidized inputs. How, then, to reconcile ECOWAS restrictive regime for SEZ-produced goods with AfCFTA liberal regime? For a long time, overlapping membership in African RTAs has been identified as one of the central and tenuous problems preventing them from realizing their full potential.¹³³ The DRC, which belongs to at least six RECs,¹³⁴ is often pointed out as an example of this problem.¹³⁵ As a result of this observation, AfCFTA aims to “resolv[e] the challenges of multiple and overlapping trade regimes to achieve policy coherence.”¹³⁶

One solution to the problem of overlapping and conflicting RoOs between AfCFTA and extant RECs for goods made in SEZs can be found in the AfCFTA conflict rules. It is not novel to deal with conflicts when two or more treaties concluded between the same parties relate to the same subject matter. The lex specialis and the lex posterior principles are often used in trade agreements to address conflicts. Under the lex specialis principle, priority is given to the more specific norm whenever two or more norms deal with the same subject matter. This is, for instance, the case of the AfCFTA Dispute Settlement Protocol, which states that it shall apply to all disputes under the AfCFTA Agreement “subject to such special and additional rules and procedures on dispute settlement contained in the Agreement.”¹³⁷ That provision further reiterates that if “there is a difference between the rules and procedures of [that] Protocol and the special or additional rules and procedures in the [AfCFTA] Agreement, the special or additional rules and procedures shall prevail.”¹³⁸ This technique would not prove helpful for RoOs of SEZs goods between AfCFTA and other African RECs since they are all specific norms.

On the other hand, lex posterior (derogat legi priori) signifies that “a legal rule arising after a conflicting legal rule prevails over the earlier rule to the extent of the conflict.”¹³⁹ The principle applies, for our purposes, to the provisions of conflicting treaties between the same parties,¹⁴⁰ in this case, between AfCFTA Agreement’s State Parties and RECs State Parties. Of the existing rules to solve conflicts between treaties, the Vienna Convention on the Law of Treaties (VCLT) has codified the lex posterior principle. Article 30 VCLT addresses conflicts regarding “successive treaties relating to the same subject matter”. It applies to all types of treaties regardless of the subject matter and the number of parties provided the treaties have been concluded at different times – as there must be an earlier and a later treaty for overlap and conflict to arise – and are still in force, and the parties are the same.

However, resorting to this provision to solve a conflict between two treaties is subject to Article 30(2) VCLT, which contains a subordination clause. Accordingly, “[w]hen a treaty specifies that it is subject to, or that it is not to be considered as incompatible with, an earlier or later treaty, the provisions of that other treaty prevail.” In other words, if a later treaty expressly concedes priority to an earlier treaty, Article 30 VCLT will not apply. This provision is, however, silent about a later treaty clause claiming priority over an earlier treaty, as is the case with AfCFTA Agreement (the later treaty) and existing RECs FTAs (earlier treaties).

Article 19 of the AfCFTA Agreement contains a conflict rule, which would be relevant for treating SEZ-produced goods under the AfCFTA RoOs and other (overlapping and conflicting) RoOS, such as the ECOWAS and EAC. Article 19(1) states the following:

In the event of any conflict and inconsistency between this Agreement and any regional agreement, this Agreement shall prevail to the extent of the specific inconsistency, except as otherwise provided in this Agreement.

This provision suggests that the liberal RoOs under the AfCFTA Agreement that confers originating status to SEZ-produced goods will prevail over conflicting restrictive RoOs in ECOWAS for the ECOWAS Member States that are also AfCFTA State Parties. As ECOWAS Member States have all signed the AfCFTA Agreement, this would, at first sight, seem straightforward if drafters had not included another element of complexity in the texts.

The phrase “except as otherwise provided in this Agreement” in paragraph 1 of Article 19 of the AfCFTA Agreement is given meaning in Article 19(2), which stipulates:

Notwithstanding the provisions of Paragraph 1 of this Article, 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.¹⁴¹

Paragraph 2 takes into account the fact that RECs are AfCFTA’s building blocs and that not all RECs are at the same level of integration. It aims to prevent integration backsliding if more advanced ones were to be required to slow down on their efforts for deeper integration. However, the incidence of this provision on RoOs in extant RECs is unclear and even contradictory. This provision ensures that RECs regimes (including their RoOs) function in parallel with AfCFTA rules when these RECs “have attained among themselves higher levels of regional integration” than under the AfCFTA Agreement. Article 8(2) of the AfCFTA Agreement’s Protocol on Trade in Goods repeats the same call by asking State Parties that are members of other RECs to maintain a higher level of trade liberalization among themselves and, where possible, improve on them.¹⁴² Improving on these higher levels of liberalization might entail further restriction on goods produced in SEZs as they may be deemed subsidized, which would conflict with AfCFTA RoOs concerning SEZ-produced merchandise.

The AfCFTA Agreement has erected “best practices” in the RECs as one of its core principles at the same level as MFN, national treatment, reciprocity, etc.¹⁴³ Unless one considers restrictive RoOs for goods produced in SEZs a bad practice, this would seem to form part of the RECs “acquis” that the AfCFTA Agreement aims to preserve.¹⁴⁴ While this relationship would benefit from further clarification, notably how RECs’ level of liberalization and “best practices” are incorporated by reference in the AfCFTA, the undefined term “higher levels of regional integration” is another hurdle worth clarifying in the future.

II. SEZ Schemes as Subsidies: Fiscal and Non-Fiscal Incentives

Even though some incentive measures could also be reviewed in light of the WTO Agreement on Trade-Related Investment Measures (TRIMs)¹⁴⁵ and the General Agreement on Trade in Services (GATS),¹⁴⁶ the Agreement on Subsidies and Countervailing Measures (ASCM) is the one that best applies to the incentives offered in African SEZs.

1. WTO Subsidies Rules and African SEZs

WTO rules allow Members to recourse to trade defense instruments by adopting restrictive measures in response to particular circumstances. For example, state support practices to promote access of their products to international markets that cause distortions and disrupt the normal functioning of the market of importing countries are considered unfair and predatory competitive practices in the territory of the importing countries. Generally speaking, a trade defense instrument, or trade remedy, is a tool that a state can use to protect itself in the event of these unfair trade practices or a massive increase in imports that could destabilize a domestic industry. While the granting of export subsidies, such as SEZ-induced measures, may be tolerated in certain instances, such as to support developing countries’ industrial policies, some export subsidies are only an aggressive and unfair means of promoting the development of exports. The role of WTO rules is to balance and distinguish between justifiable and prohibited measures.¹⁴⁷

Of the three main trade defense instruments – anti-dumping, safeguard, and countervailing measures – countervailing measures¹⁴⁸ respond to imports of products benefiting from undue subsidies, such as SEZs incentives. No blanket provision in the WTO ASCM addresses SEZs as such. Instead, the measures chosen by governments to promote SEZs are those caught by WTO rules. To better grasp the WTO ASCM’s relevance to incentives provided in SEZs, some conceptual clarifications are worth making. The WTO ASCM defines a subsidy as a financial contribution by a government or a public body conferring benefit to the recipient.¹⁴⁹

For the ASCM, a financial contribution occurs each time a government makes contributions with a monetary, economic or financial value.¹⁵⁰ As mentioned earlier regarding the measures taken to promote SEZ schemes, financial contributions can take many forms. They range from a direct transfer of funds (grants, loans, equity participation), potential direct transfers of funds or liabilities (loan guarantees, for example), and public revenues uncollected or forfeited receivables. A contribution may also have financial value without a direct transfer of funds, such as supplies of goods or services or purchase of goods. The ASCM Agreement, therefore, adopts a rather broad conception of the term “financial contribution.”

According to Article 1.1 of the SCM Agreement, the financial contribution presupposes that revenue “otherwise due” is foregone or not collected. This implies that government authorities have collected less revenue than they could have collected in normal circumstances. On the other hand, revenue foregone means that the government has given up a right to raise revenue that it should or could have collected.¹⁵¹ Since governments remain free to choose which transactions to tax, provided they respect WTO rules, what is otherwise due “depends on the rules of taxation that each Member […] establishes for itself.”¹⁵² It follows that one must compare the tax that a government would have otherwise collected “but for” a contested measure. Note that footnote 1 to Article 1.1(a)(1)(iii) ASCM provides an exception for “the exemption of an exported product from duties or taxes borne by the like product when destined for domestic consumption, or the remission of such duties or taxes in amounts not in excess of those which have accrued”, which the ASCM states “shall not be deemed to be a subsidy.”

Pursuant to Article 1.2 ASCM, the Agreement only covers “specific” subsidies. Specificity, as further elaborated by Article 2 ASCM, implies that a challenged subsidy must be able to expressly benefit certain enterprises rather than others.¹⁵³ So, suppose firms established in an SEZ are the only ones to benefit from a government incentive defined as a subsidy under Article 1.1 ASCM. In that case, the criterion of “specificity” will be met.¹⁵⁴ The first threshold question is to determine whether an SEZ program effectively meets the requirement to qualify as a subsidy.

The India – Export Related Measures dispute required the Panel to address the compatibility of an SEZ scheme with the ASCM. In that case, the US challenged a set of measures provided by India in its SEZ program, alleging that they amounted to export subsidies and were thus incompatible with the ASCM. The measures included the following: (i) exemption from customs duties on imports into, and exports from, an SEZ to every developer or entrepreneur; (ii) the exemption from India’s Integrated Goods and Services Tax (IGST) of all goods imported by a unit or a developer in the SEZ; and (iii) the deduction, from the corporate income tax base of an entrepreneur, of the export earnings of the entrepreneur’s SEZ Unit.¹⁵⁵ The Panel had no difficulty concluding that these measures amounted to subsidies as they constituted financial contributions.¹⁵⁶ The Panel argued that the reasons behind the creation of SEZs were immaterial so long as India forewent revenue otherwise due, thus conferring a benefit. Indeed, for the Panel, while “the promotion of exports is a key reason” behind the SEZ Scheme, other reasons included “the generation of additional economic activity, investment, and employment, and the maintenance of India’s sovereignty.”¹⁵⁷

Following the definition of a subsidy, it is worth noting that the ASCM prohibits two types of subsidies: subsidies contingent upon export performance¹⁵⁸ and subsidies contingent on the use of domestic products in preference to imported products (i.e. import substitution subsidies).¹⁵⁹ In India – Export Related Measures, the complainant argued that the subsidies under the SEZ program were contingent, in law and, in fact, on export performance. All three SEZ measures were found to meet this threshold as they were all contingent on export performance and thus prohibited.¹⁶⁰ It is worth noting that, contrary to the case of RoOs, where subsidized inputs can result in a denial of originating status to the final product,¹⁶¹ the ASCM does not consider duty exemption on raw materials and intermediate inputs in the manufacturing of the final products for export as prohibited subsidies.¹⁶² In other words, SEZ-produced goods made of subsidized inputs may still be denied originating status for preferential tariff treatment in ECOWAS and EAC, even if they may not fall afoul of the rule prohibiting export subsidies.

These rules that apply to all WTO Members come with some exceptions. To begin with, WTO Members say that they recognize that “subsidies may play an important role in economic development programs of developing country Members,”¹⁶³ thereby acknowledging the importance of subsidies, including those made in the framework of SEZ promotion, for developing countries’ industrial policies. Consequently, developing countries can, by virtue of the following provisions, continue to grant export subsidies contingent on export performance without violating WTO rules.

Article 27.2 of the SCM Agreement provides:

The prohibition of paragraph 1(a) of Article 3 shall not apply to:

(a) developing country Members referred to in Annex VII.

(b) other developing country Members for a period of eight years from the date of entry into force of the WTO Agreement, subject to compliance with the provisions in paragraph 4.

According to this provision, some developing countries benefit from a blanket exemption (Article 27.2(a)), while a phase-out is provided for other developing countries (Article 27.2 (b)). These provisions protect qualifying countries’ SEZ programs from the ASCM’s prohibition of export subsidies. In other words, if measures used to promote SEZs and attract investment were to qualify as prohibited subsidies contingent upon export, they would be shielded from inconsistency from WTO rules by Article 27 provisions.

Annex VII of the SCM Agreement, titled “Developing country Members referred to in paragraph 2(a) of Article 27”, provides:

The developing country Members not subject to the provisions of paragraph 1(a) of Article 3 under the terms of paragraph 2(a) of Article 27 are:

(a) Least-developed countries designated as such by the United Nations which are Members of the WTO.¹⁶⁴

(b) Each of the following developing countries which are Members of the WTO shall be subject to the provisions which are applicable to other developing country Members according to paragraph 2(b) of Article 27 when [gross national product] per capita has reached $1,000 per annum: Bolivia, Cameroon, Congo, Côte d’Ivoire, Dominican Republic, Egypt, Ghana, Guatemala, Guyana, India, Indonesia, Kenya, Morocco, Nicaragua, Nigeria, Pakistan, Philippines, Senegal, Sri Lanka and Zimbabwe.

Paragraph (b) of Annex VII provides that countries will graduate from this list if their Gross National Product (GNP) per capita reaches USD 1,000. It is fitting to note that this provision is a testimony that special and differential treatment is not a lifetime permit but a mere license that gives these countries time to gradually bring their domestic support programs into conformity with the ASCM. The license expires when a beneficiary’s GNP attains the designated threshold.¹⁶⁵ This is also in line with the spirit of the Enabling Clause, which stipulates that developing countries and LDCs are expected to graduate from special and differential treatment as their capacity to participate more fully in multilateral trade increases.¹⁶⁶

When that is the case, listed countries will be subject to paragraph 2(b) of Article 27 ASCM and required to phase out export subsidies within eight years “from the date of entry into force of the WTO Agreement.” Contrary to some developing countries’ views, especially those listed in Annex VII(b) ASCM, the Panel in India – Export Related Measures clarified that this period ran from the entry into force of the WTO Agreement in 1995 and consequently expired in 2003.¹⁶⁷ India had argued, supported by Egypt and Sri Lanka, that the period of phasing out export subsidies for graduating countries should be counted from the graduation date.¹⁶⁸ According to India, holding the entry into force of the WTO Agreement as the starting point would deprive the mandatory language of paragraph (b) of Annex VII of its effectiveness and diminish the value of special and differential treatment in favor of developing countries by distinguishing between graduating developing countries and the others.

This case proves that SEZ incentives can generate disputes and effectively go against the WTO rulebook. Indeed, the Panel found that the impugned measures, i.e. (i) exemption from customs duties on imports into or export from the SEZ, (ii) the exemption from IGST on imports into the SEZ, and (iii) the deduction of export earnings from the taxable base for corporate income tax qualified as a “financial contribution” within the meaning of Article 1 ASCM. Since they were contingent in law upon export performance – i.e. export was the condition for granting these financial contributions – the Panel had little difficulty concluding that they were incompatible with Article 3.1(a) ASCM.¹⁶⁹ The respondent was also found to either grant or maintain subsidies against the provision of Article 3.2 ASCM.¹⁷⁰

However, one peculiarity of the India – Export Related Measures dispute for African SEZs is that India had graduated from Annex VII(b) and Article 27.2(a) ASCM. Had this not been the case, its measures would have been compliant with the rules thanks to its status as a developing country. This implies that SEZ schemes from countries such as Senegal, Tanzania or Lesotho are covered by the exception and not inconsistent with the WTO ASCM. Of the 164 WTO Members, as of the date of this writing, 44 are from Africa. Of these 44 countries, 26 are LDCs.¹⁷¹ Those qualifying African countries, therefore, i.e. those that have not graduated from Annex VII(b) ASCM, continue to benefit from special and differential treatment for export subsidy rules, including subsidies provided in their SEZ programs.¹⁷² Conversely, this dispute clamors for the end in the near future of fiscal incentives that countries such as Kenya or Cameroon (featured in Annex VII(b)) use to champion their SEZ programs.¹⁷³

2. AfCFTA Subsidies Rules and their Relevance on SEZs Measures

Apart from WTO rules, RTAs frequently contain trade defense instruments to help fight against unfair trade practices. Like the WTO, RTAs also allow recourse to trade remedies when confronted with unfair trade practices. An unfair practice, such as subsidies to goods produced in SEZs, has an ambivalent effect in practice. While systematically penalizing producers of the imported goods, as they would find it difficult to compete with goods obtained under “unfair” terms and conditions, this situation could, at the same time, favor consumers, for they would purchase the same commodity at a lower price. However, governments feel the urge to intervene to ensure equal chances for all competitors. In this intervention, RTAs State Parties rank the community interests higher than consumers’ opportunity for cheaper goods against the producers’ loss in revenue along the way. Trade remedies, therefore, ensure equitable market shares among all economic actors by rebalancing rights and obligations under the RTA. Like with the RoOs, the effect of the “penalty” for subsidized goods originating in SEZs is a denial of the bound or applied preferential MFN tariffs rate.

Despite the presence of trade remedies instruments in several intra-African trade agreements,¹⁷⁴ practice does not display their use to be an everyday occurrence. The first intra-African trade dispute on the imposition of a trade remedy, which could have been resolved before a regional court but found its way to the WTO,¹⁷⁵ is a testimony of this dearth.¹⁷⁶ Nevertheless, the AfCFTA Agreement does not derogate from this established practice in almost all RTAs worldwide by providing its own rebalancing mechanism. Pursuant to Article 17(1) of the AfCFTA Agreement’s Protocol on Trade in Goods, “nothing […] shall prevent State Parties from applying […] countervailing measures.” Hence, State Parties can use countervailing measures if they determine that a subsidized product is entering their domestic market. The only requirement is to do so “subject to the provisions” of the Protocol itself.¹⁷⁷ For the practical modalities of application, this provision refers to Annex 9 of the AfCFTA Agreement on Trade Remedies while at the same time insisting on the compatibility of AfCFTA countervailing measures with the relevant WTO law.¹⁷⁸

Indeed, Article 2 of Annex 9 stipulates the following:

State Parties may, with respect to goods traded under the provisions of this Annex, apply […] countervailing […] measures as provided for in [Article 17] of the Protocol on Trade in Goods, this Annex and the AfCFTA Guidelines in accordance with relevant WTO Agreements.

Annex 9 contains procedural and substantive provisions. From a procedural standpoint, the imposing State Party must investigate before taking action. There is also an invitation to hold consultations and favor the peaceful resolution of possible conflicts.¹⁷⁹ From a substantive standpoint, the incorporation of WTO Agreements by reference signifies that AfCFTA’s countervailing measures must be WTO-compliant. AfCFTA’s countervailing duties rules thus appear to concede superiority to WTO rules on subsidies, which, for interpretative purposes, will take priority in case of conflict.¹⁸⁰ It follows that while AfCFTA’s rules have direct applicability to SEZs created by State Parties, they must still follow the WTO ASCM since the latter remains applicable in the event of a conflict.

The AfCFTA Guidelines on Implementation of Trade Remedies confirm the pervasiveness of WTO provisions by stating that “relevant provisions of the WTO Agreements […] relating to trade remedies may apply, where applicable” pending the adoption of the Guidelines.¹⁸¹ The caution—“where applicable”—and the option—“may apply”—do not really matter for our purpose since the Guidelines must still conform with “relevant WTO Agreements.”¹⁸² The caveat merely signals that AfCFTA State Parties, also WTO Members, may directly apply relevant WTO provisions on subsidies and countervailing measures to other African countries’ SEZ-made products under the AfCFTA Agreement. Whether any dispute arising from applying such a measure would be a WTO dispute or an AfCFTA dispute remains to be refined, as it could be a potential incentive for forum shopping.¹⁸³

The requirement that the AfCFTA Guidelines on subsidies must comply with WTO ASCM raises another legal issue worth clarifying in the future. In fact, as mentioned above, Article 27 ASCM provides special and differential treatment to developing countries in recognition of the important role that subsidies play in the economic development programs of these countries.¹⁸⁴ Likewise, the AfCFTA Agreement contains its own provisions to guarantee a variable geometry and special and differential treatment to State Parties to “ensur[e] comprehensive and mutually beneficial trade in goods” as a recognition that State Parties are at “different levels of economic development” or “have individual specificities.” ¹⁸⁵ Flexibilities can take the form of “an additional transition period in the implementation” of the AfCFTA Agreement.¹⁸⁶ How, then, do AfCFTA State Parties ensure compliance with the WTO ASCM while granting flexibility to deserving AfCFTA State Parties to pursue their industrial policies through SEZs? It is unclear whether granting flexibilities in the AfCFTA to State Parties in addition to special and differential treatment of Article 27 ASCM would fail to conform with the WTO rules incorporated by reference. In other words, it is not clear whether AfCFTA State Parties can get more special and differential treatment by combining the WTO and AfCFTA Agreement or whether any AfCFTA-compliant special and differential treatment must stop at the level authorized by the ASCM. The AfCFTA Trade Remedies Guidelines would need to clarify this scenario that risks diminishing the value of variable geometry and AfCFTA flexibility provisions if the WTO Agreement were to take precedence.

Since the AfCFTA Guidelines on trade remedies’ implementation remain a work in progress, its relationship with existing RECs is also imprecise.¹⁸⁷ In effect, “[p]ending the adoption of the AfCFTA Guidelines, the relevant provisions of […] regional economic communities agreements relating to trade remedies may apply, where applicable.”¹⁸⁸ This provision further consecrates the co-existence of AfCFTA rules and RECs, as alluded to earlier in the case of RoOs. Of course, one may also argue that since AfCFTA’s countervailing measures are optional – as indicated by the word “may” – State Parties may still use their relevant RECs’ provisions where they exist to tackle unfair trade practices as SEZ measures. This argument is reinforced by our discussion on AfCFTA’s conflict rule that encourages RECs that have achieved a higher degree of integration to keep it and even improve on it if possible.¹⁸⁹ It is very likely that trade remedies in these settings might have helped achieve that higher degree of integration. This implies that SEZ-produced goods of other AfCFTA State Parties, not Members of a particular REC, may be regulated by trade remedies provisions of that REC pending the adoption of AfCFTA Guidelines. Despite their recent nature, it is pretty astonishing that AfCFTA Agreement’s Annex 9 on Trade Remedies and the provisions relating to subsidies and countervailing measures are, as of date, summary, incomplete and insufficient to create a robust legal regime for goods produced in SEZs.

Interestingly, unlike WTO ASCM and some extant African RECs, Annex 9 of the AfCFTA Agreement’s Protocol on Trade in Goods does not define the term “subsidy.” However, Annex 5, which concerns AfCFTA’s rulebook for eliminating non-tariff barriers (NTBs), considers subsidies as a “category” of NTBs.¹⁹⁰ So, “government aids, including subsidies and tax benefits,”¹⁹¹ are a class of NTBs, which the AfCFTA Agreement calls for the “progressive” elimination.¹⁹² Consequently, SEZ incentives are regulated by AfCFTA’s trade remedies provisions (Annex 9 to the Protocol on Trade in Goods) and those on NTBs (Annex 5 to the Protocol on Trade in Goods). Admittedly, classifying SEZs incentive measures as NTBs (merely subject to “progressive” elimination) is a lower threshold compared to them being labelled as subsidies which are prohibited or actionable by WTO Agreements and the AfCFTA Annex 9 (incorporating WTO-covered agreements on the matter).

It is worth noting that the reference to WTO rules on subsidies in African RECs is not new. Like Article 2 of Annex 9 of the AfCFTA Agreement’s Protocol on Trade in Goods, the SADC Trade Protocol already made way for WTO law to apply when levying countervailing duties.¹⁹³ The SADC regime also resembles the AfCFTA in its treatment of subsidies as a form of NTB. When viewed as an NTB, a SADC Member can request a grace period from the Committee of trade ministers to maintain a subsidy program.¹⁹⁴ This could well be SEZ-induced measures. The main difference between the SADC Trade Protocol and the AfCFTA Protocol on Trade in Goods is that the former formally prohibits subsidies, subject to exceptions, while the latter clearly does not. The result is not far from being the same, in any case. In the case of SADC, a Member can maintain a subsidy (regarded as an NTB) and even introduce “a new subsidy” (i.e. after the entry into force of the SADC Treaty prohibiting subsidies) if the subsidy scheme conforms with WTO provisions.¹⁹⁵ This is precisely what the AfCFTA subsidies rules provide in substance. The influence of SADC Members in drafting AfCFTA trade remedy instruments, notably subsidies rules, is felt quite strongly.

Although international trade law rules in this paper have focused mainly on “products” fabricated in SEZs, and the accompanying incentives to firms, it is worth noting that the AfCFTA subsidies rules in trade in services unambiguously encourage State Parties to use subsidies “in relation to their development programmes [sic]” as “[n]othing in [the] Protocol [on Trade in Services] shall be construed to prevent” them from doing so.¹⁹⁶ This implies that, for our purpose, establishing a services-only SEZ would comply with AfCFTA rules regardless of the types of incentives granted to established firms. Also, discriminating against foreign services and service suppliers under the AfCFTA Protocol on Trade in Services may escape the disciplines, primarily because most rules depend on the extent of State Parties’ specific commitments. The non-prohibition of industrial services subsidies by the AfCFTA services rules is anything but logical. Preventing it would be useless as State Parties may exclude from market access and national treatment commitments the sectors of interest in which they use discriminatory subsidies. One may argue that State Parties expressly allowed services subsidies simply to nudge AfCFTA State Parties to open their sectors progressively after growing national champions and avoid the overuse of MFN exemption lists.¹⁹⁷ The slow WTO negotiations on subsidies in the GATS have not been helpful in framing rules on the matter, leaving each WTO Member’s policy space in this field untouched. Why would African countries feel the urge to prevent something that is not yet subject to any multilateral rule in a context where their share in global services trade is not sizable?¹⁹⁸

Available data from WTO Members’ trade policy reviews indicate a prevalence of subsidies in six sectors: tourism and travel-related services; transport services; financial services; telecommunication services; and software development services, information and communication technologies related services (ICT), data processing services and telephone call center services.¹⁹⁹ The measures include direct grants, tax incentives, preferential credits and guarantees, and equity injections.²⁰⁰ It is not always easy to distinguish between subsidies for goods and those directed at services in SEZs. Nevertheless, the WTO Secretariat notes that service providers usually benefit from SEZ incentives if they either supply their services under mode 1 (otherwise known as cross-border supply), i.e. to consumers abroad, or directly to established companies in SEZs (i.e. under mode 3 or mode 4).²⁰¹ These services generally include inspection, certification, marketing, distribution, transportation, packaging, and storage services.²⁰² Apart from these services connected to the production of goods in SEZs, literature also documents instances of services-only SEZs as a new trend among developing countries.²⁰³ It follows that services-only SEZs could also be areas that African countries may want to explore in attracting investment, including e-commerce platforms, IT-related services, call centers and fintech industries, notably for countries where manufacturing may not always be the best option.

D. Conclusion: Cutting the Baby in Half or Finding a Middle Ground on the Treatment of SEZ-Made Goods?

African countries have embarked rather enthusiastically on the path to establishing SEZs as a necessary tool for their industrialization. African SEZs pursue carefully articulated strategies combining, through a set of fiscal and non-fiscal incentives, FDI attraction in sectors as varied as agriculture, textiles, pharmaceutical and automobiles and job creation. These unilateral policies are without challenges to regional and global trade rules binding on African states. As noted in this paper, domestic laws treat the products obtained in SEZs as outside their Customs Territories, meaning, since they are destined for export, they are not directly in competition with like domestic products until they have been duly imported. However, as further discussed, there may be issues with the treatment of SEZ-produced goods in the framework of an RTA.

These challenges stem from the treatment of goods produced in these zones that compete “unfairly” with like products manufactured outside them, not only in the customs territory of the country establishing the SEZ but also those originating in the customs territory of partner countries in an RTA. In other words, while goods produced in a Rwandan SEZ may be exported to Kenyan domestic markets, Kenyan SEZ-produced goods may equally be exported to Rwanda and enter the Rwandan Customs Territory. The law must therefore provide for criteria to avoid these goods obtained under unfair conditions to compete with other goods in a region governed by a free trade agreement. In the example above, this trade agreement could be the EAC Customs Union or the AfCFTA Agreement.

From the above discussion, the dilemma confronting African countries is the following. Excluding products manufactured in SEZs from the tariff preferences could reduce the competitiveness of these goods and inputs at the continental level. However, conferring originating status to these products also comes with the risk of subjecting national production to unfair competition vis-à-vis these products considering various incentives and the lack of transparency on the origin of inputs and the nature of the operations undergone by the products exported from SEZs. While the AfCFTA currently subscribes to the prevalent RECs practice that confers originating status to SEZ-manufactured products for preferential tariff treatment, this paradigm sometimes conflicts with other RECs’ views. Considering that RECs are AfCFTA’s building blocs, which have sometimes achieved among themselves a higher degree of integration than the AfCFTA regime, conflicts of application are clearly in sight.

While considering goods produced in AfCFTA State Parties SEZs as though they originate in their customs territory is understandable in light of AfCFTA’s objectives to accelerate continental industrialization, the weakness of AfCFTA’s subsidies regime, which incorporates WTO rules on subsidies by reference, deserves urgent refinement. The imperative of guarding against unfair trade practices of AfCFTA State Parties that can distort competition at the continental level is as crucial as the need for industrial policies through tolerated SEZ incentive schemes. From a WTO perspective, several of these schemes would likely escape the subsidy rules thanks to the special and differential treatment provisions for they are developing countries. It remains unclear whether the AfCFTA special and differential treatment can be combined with WTO flexibility rules. However, as seen in the India – Export Related Measures dispute, SEZ incentives can violate WTO ASCM. Likewise, graduating countries of the WTO ASCM Annex VII, which features some African non-LDCs, could, in the future, see their measures fall afoul of WTO subsidy rules.