Kōrero / Custom before the Privy Council

Custom before the Privy Council: what five Cook Islands appeals decide, and what they do not

Ross Holmes. Kōrero, 26 October 2026. The five judgments discussed are reported in full, with headnotes, at the Privy Council reports; the Native Appellate Court decisions of 1948 to 1970, with the evidence given at each hearing, at the Appellate Minute Book reports.

Since 2012 the Judicial Committee of the Privy Council has decided five appeals from the Cook Islands. Four of them turn on customary land and custom: Descendants of Utanga and Arerangi Tumu v Descendants of Iopu Tumu [2012] UKPC 34, Baudinet v Tavioni [2012] UKPC 35, Browne v Munokoa [2018] UKPC 18 and Framhein v Attorney General of the Cook Islands [2022] UKPC 4. The fifth, Arorangi Timberland Ltd v Minister of the Cook Islands National Superannuation Fund [2016] UKPC 32, is a constitutional case on the right to property, and it matters here for one thing only: it settled that constitutional questions in the Cook Islands are decided by a proper proportionality analysis and not by the “arbitrariness” test the Court of Appeal had taken from Clarke v Karika in 1985. All five are now reported in full, with headnotes, in the UKPC series on this site.

This article is about what the four custom cases decide, and about a question they raise but do not answer: how far a Cook Islands court today may rely on what earlier Cook Islands courts said the custom was.

1. Custom is the law of the land, and a court may know it

The oldest habit of the English common law, when it met the custom of a conquered or ceded people, was to treat that custom as a local exception to the general law, to be proved as a fact by the party who relied on it. Browne v Munokoa put that habit aside for the Cook Islands. Lord Sumption, for the Board, said that native custom concerning land tenure and succession “is not a derogation from the law of the land. Subject to statute, it is the law of the land”, that courts take judicial notice of their own law, and that the need for evidence of custom is practical rather than legal: the custom is uncodified, it varies between islands and tribes, and the judges are not indigenous (para 16).

Two things follow, and the Board drew both. The first is that the decisions of the Land Court over its century are the starting point for finding the custom, because they carry the court’s accumulated knowledge and because “some stability and consistency in the matter of land title and inheritance is indispensable” (para 17). The second is that custom “is not immutable”: it has developed, and been changed, by the legislation of 1901 and 1915 and by the decisions of the courts themselves, so that an account of custom as it stood in 1894 is not an account of the custom now (paras 17, 41). On the point actually in issue, the succession of an adopted child not of the blood, the Board found the custom settled by Chief Judge Morgan’s 1968 decision in the first Emma case and the cases that followed it, and held that the House of Ariki and Koutu Nui papers of 1970 and 1977, which would have restored the custom of 1894, were proposals for law reform and not statements of the law (paras 28, 40–42).

The Board opened that judgment by recalling what it had said in Tumu and Baudinet: that ancestral land has “a special character and importance” to the indigenous people of the Cook Islands “which transcends any commercial significance”, and, in Lady Hale’s words, that the relationship of the people to their land through tikanga “is an essential component of their identity” (para 5). That is the footing for everything else in the four cases.

2. Custom yields to an enactment, but only to the extent of the inconsistency

Article 66A(3) of the Constitution, inserted in 1995, gives custom and usage effect “as part of the law of the Cook Islands”, but not where, and not to the extent that, a custom is inconsistent with the Constitution or an enactment. Framhein is the first case in which the Board had to apply that limit.

The Aronga Mana of Rarotonga, Aitutaki and Atiu had signed statements of their custom as tiaki over the kai moana, exercised through taporoporo, raui and the prohibition of destructive methods of fishing, a role extending “mei te Maunga ki te Moana”. The courts below had declined to treat those statements as the opinion of the Aronga Mana under article 66A(4). The Board held they were wrong: the statements were the opinion of the Aronga Mana, and article 66A(4) does not stop a court deciding what the custom is; it stops the court questioning the custom, once identified, for reasons outside it (paras 143–146). But the Board then held that the custom so established gave a right to be consulted only as part of the Aronga Mana’s role as guardians of the fishery, and that the Marine Resources Act 2005 had given the Ministry the principal authority over the fishery waters. The customary role was therefore inconsistent with the Act, and to that extent was no longer part of the law (paras 147–151).

The lesson is a precise one. The Board did not say the custom was wrong, or outdated, or unproved. It accepted it on the word of the Aronga Mana. It held that Parliament had legislated across it. Where Parliament has not legislated, as it has not for most of the law of customary land, article 66A(3) leaves the custom in force.

3. Time does not cure what the land law forbids

In a Western property system time heals titles. Limitation, laches and prescription exist to quiet old disputes and protect those who have relied on the record. Tumu holds that in the Cook Islands time works differently, because the statute makes it so. Native freehold land is inalienable; nothing passes by prescription; partitions and exchanges are supervised by the court; the whole scheme was built, in the Board’s words, to protect the indigenous people against exploitation by their chiefs or by Europeans (paras 60–61).

In 1912 Judge MacCormick, sitting alone, had struck five names out of Chief Judge Gudgeon’s 1905 orders for three lands at Takuvaine, “by consent”, on the application of the head of the family. Ninety-seven years later the courts of the Cook Islands validated what he had done under section 416 of the Cook Islands Act 1915, treating consent, the lapse of time and the rights of third parties as reasons to let the record stand. The Board set the validation aside. The 1912 order was made without jurisdiction; consent could not confer jurisdiction or override a protective land law; and lapse of time carried little weight where the land could not be sold or acquired by occupation and where the appellants had borne the burden of a century of archival research. What allowed the Board to restore a 1905 title after 97 years without unfairness was the appellants’ undertaking to respect the existing occupations of those who had built and lived on the land in good faith (paras 60–63).

4. Where the record is ambiguous, finality still counts

Baudinet is the counterweight. The land was Tuarea Nui section 40 in Takitumu, awarded to Makea Nui Takau on her uncontested claim of 3 June 1903. The clerk’s minute carried the words “A life interest only no power of devise”, crossed out; the sealed order omitted them. The Ngāti Raina, evicted from the land in 1932, sought to have the words restored under the slip rule, so that the whole title could be reopened.

The Board decided two points of law in the appellant’s favour. The general slip rule in section 44 of the Judicature Act 1980-81 applies to orders on investigation of title; and the failed applications of 1937 and 1950 did not bar the claim, because neither was an adjudication on the title by a court with jurisdiction to make one (paras 20–29). But a majority of three held that no slip had been proved: alterations in the 1903 minute book were not initialled as a matter of practice, the life-interest words were an afterthought written into eight Makea entries and crossed out of this one, and the most obvious explanation was a deliberate decision that they did not belong (paras 36–42). And the majority added that, even had a slip been shown, it would not have exercised the discretion a century later: Chief Judge Gudgeon used the life-interest words to protect the members of the Makea family against their own Ariki, so that family land was not turned into title land; they were never a device for handing the title to a different lineage (paras 40, 43).

Lady Hale and Lord Walker would have found a slip. Chief Judge Gudgeon had said in his reports of 1904 and 1906 that a life interest was the highest title the Māori of Polynesia had known, that an ariki was no more than a trustee for the tribe, and that in every case where an ariki claimed he had thought it right to award a life interest; every other claim by Makea Takau that day was so limited (paras 52–55). And under section 422 of the 1915 Act, which requires title to be determined by ancient custom, a custom that allowed neither mata’iapo nor ariki to alienate the land from the family, the passage of time “has much less significance than it would have in the United Kingdom” (para 62).

5. What the four cases do not decide

None of the four cases decides what the custom of any island is on any point of land tenure or chiefly title. Browne decided a question on which the Land Court had ruled consistently for fifty years, and the Board took its custom from those rulings. Framhein took its custom from the Aronga Mana. Tumu and Baudinet were about what a judge did in 1905 and 1903, not about what the custom was. So the four cases leave open the question that matters most to a Cook Islander going to court today: when a Cook Islands court has declared a custom in the past, is the court now bound by that declaration?

The materials for an answer are in the cases themselves, and in the law of New Zealand to which the Cook Islands courts have always looked.

First, Browne makes the Land Court’s decisions the starting point, not the end. The Board gave the reason: the decisions carry the court’s experience and the records of a century of successions. A decision carries that weight only if it was made on evidence of the custom. A decision made without inquiry carries none of it, and the Board’s own description of the early records, “manuscript minutes noting the decision and giving reasons in summary, not to say laconic terms” (para 18), is a warning about how much of the early case law was made in that way.

Second, the Cook Islands courts have said so themselves. In 1948 the Native Appellate Court held that native customs “are not fixed and immutable” and that “the question of whether a custom exists at a given time is a matter for the Native Land Court after proper inquiry to determine” (Re Vaimaanga section 6 and Tikioki section 45, Takitumu (1948) 1 AMB 160). In 1949 the same court found that no determination defining the customs and usages of the Rarotonga people as to land titles had ever been made and that “no full enquiry appears to have been made to establish from evidence of the inhabitants what their customs and usages are” (Re Tereora section 106B, Nikao, Avarua (1949) 1 AMB 266). The Court of Appeal in 2016 said that a court finds custom “on the basis of the evidence presented to it: the Court cannot simply make up custom out of thin air” (Hunt v De Miguel [2016] CKCA 2, para 55), and that under article 66A custom takes precedence unless ousted by statute (para 56).

Third, the common law has moved. In Attorney-General v Ngati Apa [2003] 3 NZLR 643 the New Zealand Court of Appeal held that the transfer of sovereignty did not affect customary property, that the content of a customary interest “is a question of fact discoverable, if necessary, by evidence”, and that no presumption runs against it; and it overruled its own forty-year-old decision on the Ninety-Mile Beach because the premise of that decision could not be sustained. In Takamore v Clarke [2012] NZSC 116 the Supreme Court held that the common law “has always been seen as amenable to development to take account of custom” and that tikanga must be considered wherever it forms part of a person’s heritage. In Trans-Tasman Resources Ltd v Taranaki-Whanganui Conservation Board [2021] NZSC 127 the Supreme Court treated tikanga as law that a statutory decision-maker was bound to consider. And in Ellis v R (Continuance) [2022] NZSC 114 the Supreme Court held that tikanga has been and will continue to be recognised in the development of the common law of New Zealand, that it is to be considered where relevant, and that the courts must take care not to distort it in the process; the court convened a wānanga of experts to tell it what the tikanga was before it decided how the tikanga applied. Each of those cases treats custom as something the court must find from those who hold it, and each treats an earlier judicial statement of custom as open to correction when the history shows it was wrong.

The doctrine of precedent does not stand against that. Precedent binds a later court to the ratio of a decision made on the matters tried before the earlier court. It does not turn an opinion on a question the earlier court did not try into a finding of fact, and it does not bind a court to repeat a statement of custom that was made without evidence of the custom. Article 66A(3) adds the constitutional reason: what has effect as law is the custom, not the judgment about it. A rule the courts applied before 1995 does not survive article 66A(3) because it was applied; it survives if, and to the extent that, it is the custom.

6. Why this matters: the danger in the old decisions

The point is not academic. Between 1903 and 1950 the Land Titles Court, the Native Land Court and the Native Appellate Court made thousands of orders on investigation of title, succession, partition and chiefly titles. A few of those decisions, especially on succession to ariki titles, have been cited and applied ever since as statements of the custom. The Appellate Minute Books now published on this site let anyone read those decisions in full, with the evidence given at the hearings, and the reading shows how many of them were made. Some rest on no evidence of custom at all. Some rest on a single admission by one witness. Some import the custom of New Zealand Māori, or a general account of Polynesia, or a colonial ordinance, in place of the custom of the island concerned. The Resident Commissioner of 1941, answering a petition that named six ariki who had succeeded otherwise than by primogeniture, agreed that they had, and maintained his judgment on the ground that “there is only one custom for Polynesians”.

A reader of this site, and a reader of the first three volumes of Southern Cook Islands Customary Law, History and Society, should therefore read every historical decision on custom with two questions in mind. What evidence of the custom did the court hear? And did the court find the custom of this island, or assume it from somewhere else? Where the answers are “none” and “somewhere else”, the decision is a historical fact about what the court did; it is not a reliable statement of the customary law, and the Privy Council’s own cases, read with the New Zealand cases, give a Cook Islands court every reason to inquire afresh. The volumes and the law reports on this site are written to make that inquiry possible: they put the decisions, the evidence and the sources side by side so that the question can be answered from the record rather than from repetition.

Ross Holmes LLB (Hons) is the Director of Ross Holmes Virtual Lawyers Limited and a solicitor of the High Courts of the Cook Islands and New Zealand. He appeared for the appellants in Tumu v Tumu and Baudinet v Tavioni. The five Privy Council judgments, with headnotes, are at cookislandscustomarylaw.com/ukpc; the Native Appellate Court decisions of 1948 to 1970, with the evidence, at cookislandscustomarylaw.com/amb.