China’s Zero-Tariff Policy for Africa: Stretching the Enabling Clause?

On 1^st May 2026, China implemented a zero-tariff policy covering all 53 African countries with which it maintains diplomatic relations, requiring no reciprocity. The measure builds on an earlier step: since 1 December 2024, China has eliminated tariffs on 100% of tariff lines for 33 least developed countries (LDCs) in Africa on the same diplomatic-recognition condition, meaning these countries can now ship products to China without paying customs duties at all. The policy sits at the intersection of development preferences, South-South cooperation, and shifting geopolitics, at a moment when African economies face renewed uncertainty in their trade relations with the United States and Europe.

This post examines the WTO-law basis for the measure, the harder questions raised by its extension to non-LDC African states, and the comparison with EU and US preferential schemes. It closes with a political-economy reflection on path dependency in African development against the backdrop of trade tensions.

Trade law basis and questions

We can break down China’s policy into two parts based on its trade law basis: preferences for 33 LDCs and for 20 non-LDC developing countries.

For the 33 LDCs, China relies on Paragraph 2(d) of the 1979 Differential and More Favourable Treatment Reciprocity and Fuller Participation of Developing Countries (“Enabling Clause”), which allows members to give special treatment to the least developed among developing countries. A brief overview of the 4 categories of permitted preferential treatment under paragraph 2 is reviewed here, as the categorisation is relevant for the analysis.

Paragraph 2(a) – “Preferential tariff treatment accorded by developed contracting parties to products originating in developing countries in accordance with the Generalized System of Preferences (GSP).”

Paragraph 2(b) – Differential and more favourable treatment with respect to non-tariff measures governed by GATT instruments. (*This type is not relevant to the blog discussion as it is about non-tariff measures).

Paragraph 2(c)- Regional or global arrangements among developing countries for the mutual reduction or elimination of tariffs.

Paragraph 2(d) – Special treatment of LDCs.

In general, the Enabling Clause is an exception to one of the core pillars of the multilateral trade system, namely, the Most-Favoured-Nation (MFN) principle in Article I:1 of GATT 1994. Under the Enabling Clause, Paragraph 2(d) permits a Member to single out LDCs as a category for deeper preferences. Moreover, the 1999 WTO LDC tariff waiver also allows preferential treatment for least-developed countries, which was extended until 30 June 2029, allowing developing-country Members to grant preferential tariffs to UN-designated LDCs without extending them to all WTO Members. The waiver has further notification requirements and that the treatment be generalized, non-reciprocal and non-discriminatory. China formally notified the WTO of its expanded zero-tariff policy for LDCs in June 2025 to comply with the transparency mechanism.

For the 20 non-LDC African countries, the legal questions are trickier. On a plain reading, the Enabling Clause, paragraph 2(a) permits preferences for developing countries generally, but only on a non-discriminatory basis among similarly situated beneficiaries, a point clarified by the Appellate Body in EC-Tariff Preferences 2004 (para.173). Granting zero tariffs to African developing countries while excluding non-African developing countries at comparable levels of development risks inconsistency with this requirement. Yet whether we can place the treatment of 20 non-LDCs under which categorisation under paragraph 2 is not as clean as it first appears. At least two wrinkles deserve attention.

The first concerns China’s framing. China appears to be addressing the non-discrimination problem by framing the arrangement as a transitional, two-year measure pending the conclusion of the China-Africa Economic Partnership for Shared Development. Chinese official describes the zero-tariff measure as “South-South cooperation” rather than as GSP under paragraph 2(a). South-South cooperation is the conceptual frame that fits the developing-country-to-developing-country flow, and it leans on paragraph 2(c) rather than 2(a). But to be noted, paragraph 2(c) on its own terms covers mutual reduction or elimination of tariffs, which is also not a proper niche, either, since there is no reciprocal requirement for African countries. If and when the China-Africa partnership is concluded as a free trade agreement, the legal basis would shift to GATT Article XXIV (regional trade agreements), which permits reciprocal liberalisation but requires “substantially all the trade” to be covered and the agreement to be notified to the WTO. The interim 2-year period for the future agreement, therefore, raises separate WTO-compatibility questions.

The second concerns the granting state. If the Chinese measure is treated as functionally equivalent to a GSP, paragraph 2(a) imposes a textual constraint: the preferences must be “accorded by developed contracting parties”. Unlike paragraph 2(d), which is silent on the granting state, the developed-country qualifier in paragraph 2(a) is not accidental. The 1979 drafters worked within a framework in which the developed/developing categories were treated as relatively stable and preferences flowed in one direction. The WTO adopts a self-declaration system for dividing developing and developed countries, and China has considered itself a developing country, considering the overall nationwide development status, despite criticism from its counterparts. However, at the annual meeting of the UN General Assembly in September 2025, Chinese Premier Li Qiang announced that China would no longer seek new special and differential treatment (SDT) for developing countries in the WTO, in current or future WTO negotiations. The announcement is narrower than some media reports suggested: e.g. “China gives up the ‘developing country’ status”. Instead, Premier Li emphasised that “this does not involve any change to China’s status as a developing country, whether within the WTO framework or in any other context.” Developing-country status in WTO law is not a single switch but a bundle of entitlements; the announcement targets only one of these, namely the active claiming of new SDT, and only prospectively. As a unilateral act under international law, China is bound only by the scope of legal obligations it expressed. Paragraph 2(a) accordingly remains textually inapt, both because of the granting-state qualifier and because the measure is not, in form, a GSP.                                                       

Conditionality without calling it conditionality?

Existing preferential trade access, e.g., by the US and EU, depends on the beneficiary’s compliance with non-trade requirements. The Chinese measure also invites comparison with the existing models of trade preference conditionality.  The US African Growth and Opportunity Act (AGOA) conditions duty-free market access on certain governance, human rights, and rule-of-law benchmarks. Similarly, the EU’s GSP+ requires ratification of 27 international conventions on human rights, labour, and environmental protection. The EU’s Everything But Arms (EBA) can be suspended for serious human rights violations, for instance, as occurred with Cambodia in 2020.

The Chinese model differs sharply. It does not attach trade preferences to governance reform, human rights conventions, or domestic regulatory benchmarks. This is consistent with China’s long-standing emphasis on sovereign equality and non-interference and departs from the “conditionalities regime” embedded in Western trade preferences. Yet the absence of substantive conditionality does not mean the absence of any condition at all. The Chinese approach also has its own “condition”, which depends on the diplomatic ties (excluding Eswatini, which recognises Taiwan). The same diplomatic-recognition condition appeared in the June 2025 LDC notification and was reiterated in the May 2026 expansion.

Paragraph 2 of the Enabling Clause enumerates the categories of permitted preferences, while paragraph 3 sets conditions that any such preference must satisfy. Notably, paragraph 3(a), especially the last half of the sentence, is the relevant constraint, which explicitly requires that any differential and more favourable treatment provided shall not “raise barriers to or create undue difficulties for the trade of any other contracting parties”. Notably, Eswatini is not an LDC on the UN LDCs list. Whether the exclusion of Eswatini causes “undue difficulties for the trade” is a fact-specific question that, in principle, the excluded state could raise. In general, it illustrates how preferential schemes can carry implicit conditionality even when no formal conditions are written into the preference itself.

Not the final solution, but a window of opportunity

The policy leads to deeper political economy issues. The first is the dependence of development. We can compare with an existing scenario. Over roughly the same period but a bit earlier, the US has imposed sweeping “reciprocal tariffs” on imports, in particular affecting African states (for detailed analysis, see Olabisi D. Akinkugbe), dismantled the United States Agency for International Development (USAID) as an independent agency, frozen or terminated billions in foreign assistance programs, and allowed AGOA, the cornerstone of US-Africa trade preferences since 2000, to expire on 30 September 2025. Although AGOA has been reauthorised through 31 December 2026 with retroactive effect, we can see the vulnerability of unilateral preferences granted under the will of one State.  Despite the diplomatic goodwill and the substantial market access benefits it offers, the Chinese arrangement is structurally similar, with a unilateral grant at the discretion of the granting state, which may be modifiable or revocable at will. The deeper question, whether Africa can ever escape the precariousness of depending on great-power goodwill, regardless of which power, is not answered by simply changing the source of preferences.

Second, and related to the first, what needs closer examination is the structure of trade itself. One cannot ignore the genuine short- and medium-term gains of the zero-tariff policy, particularly for African small and medium-sized enterprises (SMEs), especially in labour-intensive sectors (See, UNCTAD Outlook 2024). Previous studies found that 89% of Africa’s exports to China are extractives, such as oil and copper, while 94% of China’s exports to Africa are manufactured goods, such as telecom equipment and fabrics. Lower tariffs do not, by themselves, change this asymmetric structure, and further, they may even entrench it by making the status quo more profitable without addressing the underlying barriers to industrialisation. The endogenous motivation for domestic industrial upgrading can be weakened as the existing comparative advantages in primary commodities are further reinforced. Structuralist development theory suggests that trade liberalisation without industrial policy can freeze economies in low‑value sectors (see, classically, Ha-Joon Chang). Trade patterns can “lock in” structural positions, then the real question is what comes next.

The most important variable is not what China does, or what the United States does, but what Africa does. In particular, we need to re-examine the scenario in the context of the African Continental Free Trade Area (AfCFTA), which has been in effect since 2021 and was conceived precisely to address the problem of dependency. AfCFTA should not be understood as a parallel issue, but as the institutional condition for making external preferences developmentally useful. Tariff-free access to China is more transformative if extractives or agricultural goods are processed along African value chains before export. As analysed by the WTO and the United Nations Economic Commission for Africa (ECA), full AfCFTA implementation could substantially expand intra-African trade, with major gains concentrated in industrial sectors.

The Chinese policy is therefore best understood as a window of opportunity, not a final solution. It raises further legal questions about the 1979 Enabling Clause, which is stretched to accommodate a major economy that grants preferences at similarly developed-country levels of generosity while remaining classified as developing, and that conditions access on diplomatic alignment rather than on internal governance requirements. Beyond legal questions, we can foresee that using the policy well or poorly can lead to divergent outcomes, either structural transformation or the substitution of one form of dependency for another. The decisive question, therefore, is whether the continent can convert unilateral preference into regional production capacity.