Rwanda's £100 Million Loss Over One Word

The One Word That Cost Rwanda £100 Million: Inside the PCA's Article 13 Ruling

A PCA tribunal held that Rwanda's reply of "acceptable" in a November 2024 note verbale amended a binding treaty — reshaping how VCLT Article 13 governs diplomatic exchanges.

Rwanda lost more than £100 million in Kigali on 15 May 2026 because its foreign ministry replied "acceptable" to a British note verbale in November 2024 — and a three-judge Permanent Court of Arbitration tribunal chaired by Peter Tomka decided that one word, read against the surrounding conduct, was enough to amend a binding treaty under Article 13 of the Vienna Convention on the Law of Treaties. The award, in Republic of Rwanda v. United Kingdom (PCA Case No. 2025-45), is the first modern inter-state ruling to squarely apply the sparsest-litigated provision of the VCLT, and its holding — that "circumstances" and "subsequent conduct" can carry consent under Article 13(b) even where the exchanged texts do not — quietly hands governments a sharper knife for unwinding awkward treaties by diplomatic correspondence.

What the tribunal actually decided

The dispute crystallised on two dates. On 13 November 2024, the Foreign, Commonwealth and Development Office sent Kigali a note verbale proposing that Rwanda forgo two £50 million payments due in April 2025 and April 2026 under the Economic Transformation and Integration Fund (ETIF). The UK note used the classic exchange-of-notes closer: "If the proposals set out above are acceptable to the Government of Rwanda, I have the honour to propose that this Note and your reply to that effect will place on record the understanding of our two Governments in this matter." Rwanda's Foreign Minister replied on 14 November 2024 that "the proposed arrangements are acceptable." The award, hosted by Aceris Law, records both texts in full.

Three months later, after London suspended most bilateral aid over Rwanda's role in the M23 offensive in eastern DRC, Kigali purported to "rescind" the November arrangements. Rwanda then filed a Notice of Arbitration in November 2025 claiming £50 million for Year 2, £10.4 million for Year 3, and £6 million (or an apology) for the UK's alleged breach of Article 19 on refugee resettlement, as reported by JURIST.

The tribunal — with Judge Joan Donoghue (former ICJ President) representing the UK and Professor Mohamed Abdel Wahab representing Rwanda — rejected every claim. On the money, the majority held that the November 2024 exchange amended paragraphs 2.3.1 and 2.3.2 of the 2024 Finance Agreement. On resettlement, it found that Article 19 imposed only a joint obligation to negotiate arrangements in good faith, and that Rwanda had never tabled a concrete resettlement proposal before raising it as a dispute in March 2025, according to the summary published by the American Society of International Law.

Why Article 13 is the real story

Domestic politics dominated coverage — Al Jazeera framed it as a warning shot to European "return hubs"; the BBC as a win for Downing Street. Both miss the point. The award is the first inter-state ruling to give sustained content to VCLT Article 13(b), the provision Polish delegate S. E. Nahlik pushed into the Convention at the 1968–69 Vienna Conference to catch consent expressed by exchange of instruments where the texts alone do not clinch it.

Article 13 provides that consent by exchanged instruments arises when either "(a) the instruments provide that their exchange shall have that effect; or (b) it is otherwise established that those States were agreed that the exchange of instruments should have that effect." The provision was not in the ILC's original draft — it was inserted at Nahlik's initiative, and it has been analysed at length in almost no modern jurisprudence. As Alexandru Bolintineanu observed in the American Journal of International Law, the Vienna text on exchange of instruments deliberately left the "decisive role" to "the will of the negotiating states" — an invitation the Tomka tribunal has now accepted.

The tribunal split on how far Article 13(b) can travel. All three arbitrators agreed that Article 13(a) did not apply: the wording of the exchanged notes was too ambiguous, and Rwanda's "acceptable" did not mirror the UK's specific proposal. The majority — Tomka and Donoghue — then held that Article 13(b) was satisfied by "the circumstances of the exchange, prior negotiations, draft history, and subsequent conduct," a framework close to the one described in the Aceris Law summary of the award. In effect, they imported the interpretive furniture of VCLT Articles 31–32 into the consent question at Article 13.

The dissent that reframes the doctrine

The most important document in the case is not the award — it is Mohamed Abdel Wahab's dissenting and separate opinion, published in full via Jus Mundi. Abdel Wahab agrees with the outcome on Articles 18 and 19 and even accepts that the Year 3 payment was not due. But he refuses to accept the majority's Article 13(b) reasoning, and his objection is doctrinal, not political.

Writing that the notes verbales lacked "clear and convincing evidence of concurrence of wills over the forgoing of the ETIF payments," Abdel Wahab argues that Article 13(b) requires a high threshold — a finding that the states unequivocally agreed the exchange itself would carry consent to be bound. The differences in wording between the UK and Rwandan notes, the "ambiguity in the use of the word 'arrangements'," and the absence of any clear Rwandan waiver of future ETIF payments meant, in his view, that consent under 13(b) was not established. The Kigali government seized on precisely this passage: as the BBC reported, Rwanda's foreign ministry noted that the dissent showed "the November 2024 exchanges relied on by the UK did not validly change the financial arrangements."

That is the split that matters for future cases. The majority's Article 13(b) is elastic and evidence-driven. Abdel Wahab's is textualist and strict. Practitioners advising ministries on the effect of exchanged notes now have to negotiate against both.

Who wins beyond the courtroom

The immediate winner is HM Treasury. According to the Home Office breakdown published after the scheme's cancellation, the UK had already transferred £290 million to Rwanda for four voluntary relocations — a per-capita cost that made the programme politically indefensible — but avoided a further £220 million in scheduled payments. The award locks in those savings by removing Rwanda's principal legal argument for clawing back the April 2025 and April 2026 tranches.

The larger winner is the technique of governing-by-note-verbale. The UK repealed the Safety of Rwanda (Asylum and Immigration) Act 2024 through Section 40 of the Border Security, Asylum and Immigration Act 2025, formally notified termination of the Asylum Partnership Agreement on 16 December 2025, and let it expire on 16 March 2026. But the money question had already been resolved four months earlier, through a two-page diplomatic exchange. That is a template. It is also a warning.

The losers are twofold. First, Kigali: Rwanda incurred what its attorney general Emmanuel Ugirashebuja called "significant costs" preparing reception capacity and gets nothing back, since the ETIF payments already made were, by the treaty's own terms, non-recoverable. Second, and less noticed, foreign ministries that use "acceptable" as a boilerplate diplomatic courtesy. The tribunal has now confirmed that in a Vienna Convention world, that word can bind — and the burden shifts to the responding state to make its reservations explicit on the face of the reply.

The second-order effect: return hubs and the EU

The award lands squarely on the desk of the European Commission, which was on the same day finalising talks on its Returns Regulation and its own "return hubs" project. As Al Jazeera noted, the ruling "bodes ill" for third-country migration arrangements. But the more precise lesson is not that such deals are legally risky in substance — it is that their unwinding is legally cheap when it is done by exchange of notes.

For the EU, Italy (whose Albania deal was scrapped), Denmark and Germany — all exploring extraterritorial processing — the ruling means the exit ramp is wider than it looked. A carefully drafted diplomatic exchange, in a moment when the receiving state is politically weakened, can now defensibly amend the financial obligations of a binding treaty without recourse to formal termination clauses. That is a governance point the wire coverage has almost entirely missed.

What to watch

• PCA jurisprudence uptake. Whether investment-treaty and inter-state tribunals cite the Tomka majority's Article 13(b) framework in the next 12–18 months. The award is unpublished in the PCA's official case list beyond the case number, but a formal publication and commentary in the American Journal of International Law is likely in ASIL's next volume; ASIL's ILIB summary already signals the direction.
• UK-Rwanda residual claims. Rwanda's justice ministry has publicly accepted the ruling but flagged the dissent. Watch whether Kigali seeks an interpretive reference or files a fresh claim over the £6 million Article 19 resettlement issue, which the tribunal declined to award because no concrete proposal had been tabled.
• EU Returns Regulation. The Commission's Returns Regulation text is expected to be finalised in the second half of 2026. Its language on the legal form of third-country arrangements — treaty, MoU, or exchange of notes — will now be drafted in the shadow of the Tomka award.
• Domestic UK politics. With Keir Starmer having stepped down (per JURIST's dispatch feed), the Conservative opposition has already argued the ruling vindicates the original policy. Whether a future government tries to revive a Rwanda-style scheme by exchange of notes rather than treaty is the specific policy question the award has now made viable.

The Bottom Line

The Tomka tribunal did not resolve a migration dispute — it resolved a treaty-law puzzle that had been sitting untested in Article 13 of the Vienna Convention for fifty-seven years. By holding that consent to amend a binding treaty can be "otherwise established" under Article 13(b) from the circumstances and subsequent conduct surrounding an ambiguous note verbale, the majority has quietly expanded the toolkit governments use to walk back inconvenient commitments. Abdel Wahab's dissent will be the citation of choice for states that want to close that door; the majority opinion will be the citation of choice for the ones that want to keep it open.