One Strong Anchor, Fifty Missing Flags: South Africa and the Convention's Thinnest Continent

Executive Summary

Lay the Hague Convention's membership map over Africa and only about a dozen of the continent's fifty-four states have ever joined — and for cases between African states the global statistics contain almost nothing. Which is exactly why South Africa matters beyond its borders: it is the continent's one deep, functioning Hague system (a 75% return rate in 2021 at a 152-day average), and its Constitutional Court decided Sonderup v Tondelli (2000), which reconciled the world's most emphatic "best interests paramount" constitution with the Convention's summary-return machinery. It did so not by carving out an exception but by engineering the return — attaching protective conditions and a mirror order so the treaty's design and the child's welfare could both be honoured. The lessons are regional: grade operation, not membership; resource anchors like South Africa as infrastructure; and, above all, count Africa. This is educational, not legal advice.

Introduction

Lay the Convention's membership map over Africa and the picture explains itself: of the continent's fifty-four states, only about a dozen have ever joined the 1980 Convention — among them South Africa, Morocco, Tunisia, Mauritius, Seychelles, Burkina Faso, Guinea, Lesotho, Zambia, Zimbabwe, Gabon (where the treaty has never become operational), and, since 2023, Botswana and Cabo Verde. For the rest — including giants like Nigeria, Egypt [article #20], Ethiopia and Kenya — there is no return mechanism at all, and for cases between African states the global statistics contain, in effect, nothing: several African members did not respond to the HCCH studies in any year, and the intra-African flow of abducted children is the least-counted corridor on Earth.

Which is precisely why the continent's one deep, functioning Hague system matters beyond its borders. South Africa joined in 1997, runs a docket that produced a 75% return rate in 2021 — among the highest recorded for any active system (9 of 12 applications; 5 voluntary, 4 judicial, zero refusals) at an average of 152 days — well above the 39% global average — and, a generation ago, its highest court decided a case that anticipated where the entire Convention world would eventually arrive.

Legal Background: paramountcy meets summary return

The tension at the heart of this article is constitutional. Section 28(2) of South Africa's Constitution provides — in a sentence with few equals anywhere — that "A child's best interests are of paramount importance in every matter concerning the child." The Hague Convention, by contrast, orders a child's summary return to the country of habitual residence, so that that country's courts decide the child's future — deliberately without a full best-interests inquiry in the returning state. (As throughout this series: a Hague return decides only the forum, not custody. Return ≠ custody.) How can a constitution that makes best interests paramount in every matter coexist with a treaty that returns children without weighing their individual best interests? South Africa answered that question a decade before Europe's courts wrestled with the same collision — and its answer runs through the Constitution's own limitation clause (section 36), which permits a right to be limited where the limitation is reasonable, justifiable and proportionate.

What happened

The case reached the Constitutional Court in late 2000 as Sonderup v Tondelli (decided 4 December 2000, per Goldstone J). A young child had been retained in South Africa by the mother after an agreed visit; the child's home, and the father, were in British Columbia, Canada. Under the Convention — domesticated in South Africa in 1997 — the retention was wrongful and return would ordinarily follow.

But the mother's argument wrote itself: a treaty that returns children summarily, without a best-interests inquiry, cannot coexist with a constitution that makes best interests paramount in every matter. It was Neulinger's dilemma [article #6] and Germany's Tiemann question [article #9], posed under one of the world's most child-centred constitutional texts.

The Constitutional Court, per Goldstone J, held the treaty and the constitution together with two moves:

1. The Convention serves best interests at the level of design — and any intrusion is a proportionate, justified limitation. Returning abducted children quickly, so that the right court decides their future, protects children as a class; the short-term intrusion on the individual child is a limitation of section 28(2) that is reasonable and justifiable under section 36 — the more so because the Convention's own Article 13 and Article 20 exceptions already mitigate it. Forum, not custody; the child's long-term interests are decided at home.
2. The returning court must protect the child through the transition — with real conditions. The Court did not merely permit protective conditions; it built them into the return order, exacting undertakings from the father addressing the child's and mother's interim care, custody, maintenance and ancillary costs on return to British Columbia — and requiring him to obtain a mirror order from the Supreme Court of British Columbia in the same terms, so the protection would be enforceable at the other end. This is the safe-return approach that, two decades later, became the global answer to the grave-risk dilemma — the HCCH's 2020 Guide, the POAM framework, Golan v. Saada [article #14]. South Africa's distinctive contribution was to root it in constitutional compulsion: a paramountcy clause obliges a court not merely to order the return but to engineer it.

The child returned to Canada under those conditions. And South African courts have carried the doctrine since — the system is visibly alive today, with High Court Hague judgments issuing steadily (including reasoned refusals where defences are made out, and returns where they are not) and the Supreme Court of Appeal ordering a Convention return as recently as 2026 [sources 4–5].

The continent around the anchor

Three facts frame Africa's Hague reality, each with a policy edge:

• Morocco is the bridge. A state whose family law is grounded in Islamic (Sharia-influenced) principles, Morocco acceded in 2010 and received 19 incoming return applications in 2021 — though, on the cases with usable timing data, it averaged 334 days, among the slowest measured [Annexes 2, 7; ¶105]. Its membership, nurtured through the HCCH's "Malta Process" dialogue between Convention and Sharia-influenced family-law systems, is standing proof that the treaty and religious family law can coexist [the article #20 question, answered affirmatively].
• Membership without machinery is common. Gabon's accession never entered into operation; several members have never responded to a statistical study; some appear in no dataset at all. An accession that produces no Central Authority, no designated courts and no data is a flag, not a system [the Mexico lesson, #11: implementation is the treaty].
• The uncounted corridors are here. Africa's cross-border families — across the Nigeria–Ghana, Kenya–Uganda, SADC and diaspora corridors — generate abduction cases that appear in no Hague statistic, no US report annex beyond a handful of countries, and no NGO dataset. The research consensus that "no comprehensive statistics exist" is nowhere truer than between African states.

What This Shows About the Limits of the Hague Convention Alone

Africa exposes two different limits at once. The first is reach: the Convention simply does not exist across most of the continent, so for the majority of African cross-border families the treaty is not a weak remedy but no remedy — and the intra-African corridor is so uncounted that the problem cannot even be sized. The second is the limit Sonderup answers: even where the Convention applies, its summary-return logic sits in visible tension with a rights-based constitution, and the treaty's text alone does not resolve that tension — a national court has to, through proportionality and protective conditions. South Africa shows both the ceiling of the Convention's map and the constitutional craft required to make it fit a modern bill of rights.

What Parents and Professionals Should Understand

For parents, the practical reality — context, not legal advice — is starkly geographic: if a child is taken to a South Africa, a Morocco or another operating member, the machinery exists and works; if taken to one of the continent's many non-members, or between two African states, there may be no return mechanism at all, and the case will turn on local law, diplomacy and the destination's own courts [article #20]. For policymakers, three lessons follow. Paramountcy clauses and summary return can coexist — Sonderup is the template: uphold the treaty's design and discharge the constitutional duty through protective engineering of the individual return, rather than Neulinger's dead end or a full-merits trial. Regional anchors should be resourced as infrastructure — South Africa's courts and jurisprudence are, functionally, the continent's Hague academy, and judicial-network expansion and accession mentoring through Pretoria (as the Malta Process worked through Rabat) is the highest-leverage growth strategy the Convention has in Africa. And accession campaigns must sell machinery, not signatures — the gap between Gabon (member, inoperative) and South Africa (member, 75% returns) is the whole story.

Limitations

South Africa's 2021 figures rest on a small caseload (12 return applications), so they indicate a pattern rather than a precise ranking, and the 2021 data was affected by the pandemic. The list of African members reflects the HCCH status table and can change; readers should check current status for any specific country pair. Sonderup is summarised from the judgment and scholarly commentary. This is educational and not a substitute for advice from a qualified lawyer in the relevant jurisdiction.

Conclusion

South Africa is a single strong flag on a map that is mostly blank. Its courts return most of the children brought before them, quickly, and its highest court showed twenty-odd years ago how a rights-based constitution can carry a summary-return treaty without breaking either — by engineering the return rather than refusing it. But one anchor cannot hold a continent. The task the numbers set is not to admire South Africa; it is to build more South Africas — and, first, to count the cases that today disappear between African states with no one to record that they happened at all.

Frequently Asked Questions

Does the Hague Abduction Convention apply across Africa? Only partly. About a dozen of Africa's 54 states are members (including South Africa, Morocco, Tunisia, Zambia, Zimbabwe and, since 2023, Botswana and Cabo Verde). Most of the continent — including Nigeria, Egypt, Ethiopia and Kenya — is outside the treaty, so there is no automatic return mechanism.

**What did Sonderup v Tondelli decide?** The South African Constitutional Court held that ordering a child's Hague return is compatible with the constitutional rule that a child's best interests are paramount (s 28(2)), because the intrusion is a proportionate, justified limitation under s 36 — provided the court attaches protective conditions. It ordered the child returned to British Columbia under undertakings and a mirror order.

How can "best interests are paramount" coexist with summary return? Through proportionality. A court treats prompt return as serving children's interests as a class, limits the individual intrusion so far as reasonable, and protects the specific child through enforceable conditions on the return — leaving the final custody decision to the home country's courts.

Why is intra-African child abduction so hard to measure? Most African states are not Convention members and do not report to the HCCH studies, and cases between African states appear in no international dataset. The result is that the size of the problem on the continent is essentially unknown.