Hiding in Plain Sight: Chief Justice Winkelmann’s revolutionary vision

LawNews

This column was first published in LawNews on 15 August 2026. It continues a series examining the Supreme Court’s departure from the constitutional limits of judicial power.

Suppose a public official holds an office with limited powers. The limits are not in doubt. They are written down, and the official has never denied them. Yet the office acts, case by case, in ways that are hard to square with the limits. Then, in a series of measured public addresses, the official sets out an account of the office under which its powers exceed those limits. The account is careful and scholarly, delivered to expert audiences. No one objects, because it is offered as observation, a reflection on how things have come to work.

This is not a hypothetical. It describes a constitutional theory that has been articulated in a series of recent addresses by one of New Zealand’s most senior public officials: the Chief Justice, Dame Helen Winkelmann. And The limits are not mere “technicalities.” They are the requirements of a free society: laws mean what they say, those who hold public power are bound by those laws, and no class of unelected experts, however able or well-intentioned, is permitted to dictate the terms on which everyone else lives.

New Zealand has, by international standards, enjoyed an unusually pure version of that settlement. There is no upper house and, unlike the United States or Canada, no written constitutional charter under which laws may be struck down by courts. Parliament is democratically elected and – as the Attorney-General has recently reminded us – is sovereign. The courts interpret and apply what Parliament has enacted – they decide cases with constitutional consequences, interpret statutes and settle the common law’s unanswered questions. And section 4 of the New Zealand Bill of Rights Act 1990 makes plain that no court may decline to apply legislation on the ground that it is inconsistent with the rights affirmed in the Bill.

The threat to this settlement was foreseen and condemned at the time of the Supreme Court’s formation. In October 2004, Deputy Prime Minister Michael Cullen addressed a public law conference on parliamentary supremacy and fundamental norms. He set out the orthodox position: for the courts to “find” that a “higher law exists which modifies the constitutional status of the New Zealand Parliament” would “amount to constitutional change by stealth.” And he named a second danger: “advocacy of constitutional change by sitting judges,” which “invites the inference that the Courts may not be impartial in their decision-making.” He did not labour them. New Zealanders, he said, had “every reason to be confident” in their Parliament and their judiciary.

Twenty years on, the two warnings have converged. The possibility Cullen described has been articulated in three extrajudicial addresses that outline a vision from the country’s current Chief Justice. Yet they have not been read together for what they are.

The theory stated

The most recent, the Chief Justice’s April 2026 George Winterton Memorial Lecture at the University of Sydney, states the vision most fully. The lecture is named for a major Australian constitutional scholar, and the audience included the Chief Justices of Australia and New South Wales. Chief Justice Winkelmann’s subject was New Zealand’s constitutional settlement. The “braided river” was her metaphor. The constitutional landscape – the constitution itself – is, she said, “constantly being reshaped by legislation, by court decisions, and by significant historical and even cultural events.”

The passage names three forces by which the constitution is reshaped. We should take each in turn.

The first force is uncontroversial. Parliament reshapes the constitutional order through legislation – a function for which it is elected. Acts of Parliament, debated in public and voted on by elected representatives, may alter the constitution in any session.

The second force is court decisions. The orthodox account does not deny the courts a hand in the constitution. What it denies is rank. The courts’ contribution is subordinate: made as ordinary law, and remade whenever Parliament chooses. The Chief Justice’s formulation drops the subordination. Court decisions are named, in parallel with legislation, as one of the forces by which the constitution is “constantly being reshaped.”

The third force – historical and cultural events – is the one a casual reader might pass over. Events can remake parts of an uncodified constitution unaided: conventions emerge from political practice. The lecture’s own example is the convention articulated during the 1984 transfer of power from the outgoing Muldoon Government to the incoming Labour Government.

But the lecture claims more for the third force than convention. The Chief Justice says that “historical and social forces have ongoing potential to shape and order New Zealand’s public law.” Public law is doctrine. Before an event can shape doctrine, some institution must decide which events count and what law they yield.

On the orthodox account, the answer to “which institution?” is “Parliament”: the legislature translates historical and cultural change into law. The Chief Justice’s account does not reserve that function to Parliament. By listing “significant historical and even cultural events” alongside legislation and court decisions, her formulation permits either Parliament or the courts to perform it. Her account, therefore, licenses the courts twice over: a court decision may reshape the constitution in its own right, and it may reshape it again as translator of the third, converting historical and cultural change into doctrine the legislature has not enacted.

Yet no statute confers this power on the courts. The Supreme Court Act 2003, which established the Court, stated a purpose, not a power – that important legal matters be resolved “with an understanding of New Zealand conditions, history, and traditions.” Of that section the Chief Justice has said she was “disappointed to see that section was not carried forward into the Senior Courts Act 2016.” What both Acts do state is that “Nothing in this Act affects New Zealand’s continuing commitment to the rule of law and the sovereignty of Parliament” – a denial that the statute disturbs a commitment older than itself, not a grant of anything.

The vision was not improvised for Sydney. Eight months earlier, at Monash, in the Lucinda Lecture “Continental Drift: Constitutional Development and Divergence in New Zealand and Australia”, the Chief Justice had put it in nearly the same words: a constitutional landscape “capable of being reshaped by legislation, by court decisions and even by significant events.” The Chief Justice enlisted Professor Claudia Geiringer in support – “we are constantly reconstituting ourselves ... just a series of constitutive iterations.” But Geiringer’s observation describes how an uncodified constitution may change over time, not who holds the authority to change it. The borrowing is accurate. The use is not.

Before that, at the Supreme Court’s 20th-anniversary conference in Auckland in February 2024, in “The Supreme Court: A View from the Top”, the Chief Justice asked “A constitutional court?” and answered it. “For my part,” she said, “it is obvious that the Supreme Court of New Zealand is a constitutional court, notwithstanding the absence of a power to invalidate legislation … indeed in part because of [its general jurisdiction].” As she put it, “The very nature of the work of the courts is constitutional.” By the time the braided river reached Sydney, the claim beneath it was two years old.

A reply considered

The natural defence is that the Chief Justice’s formulation is descriptive rather than prescriptive. On this reading, the Chief Justice is not asserting a constitutional role for the courts in reshaping the constitution. She is acknowledging that courts in any system, including New Zealand’s, contribute to the development of law over time – by interpreting statutes, applying them to new cases, and developing the common law in areas Parliament has left open. That much A.V. Dicey himself accepted. The braided river, on the descriptive reading, is an observation about how an uncodified constitution operates in practice. It is not a brief for judicial supremacy.

The descriptive reading is attractive, and the lecture lends it real support. Before the braided river, the lecture recounts Fitzgerald v Muldoon [1976] 2 NZLR 615 – with Wild CJ declaring unlawful a Prime Minister’s attempt to suspend a statute by press release – and the caretaker convention articulated after Muldoon’s 1984 refusal to devalue. A court decision and a significant event, each adding a strand to the settlement, and each impeccably orthodox: the court policed the boundary Parliament had set, and the convention bound the executive, not the legislature. If those were the account’s illustrations, there would be nothing to see. But two features of the lecture make the descriptive reading incomplete.

The first is that the lecture does not leave “court decisions” abstract. Having named them as a force that reshapes the constitution, it spends its length showing the courts doing the reshaping. Take the Treaty. The lecture says its significance “was first acknowledged in the political and legislative spheres,” and that “[i]n turn, this has led the courts to recognise the Treaty as a document of constitutional significance.” “In turn” is doing a great deal of work in that sentence. The first half is true. Parliament enacted the Treaty of Waitangi Act 1975; Parliament wrote Treaty clauses into statute after statute; the Lands case construed one of them (New Zealand Maori Council v Attorney-General [1987] 1 NZLR 641 (CA)).

But the constitutional status that the lecture describes does not rest on those clauses. It rests on a presumption of the courts’ own construction – that Treaty principles apply even where a statute is silent, unless Parliament excludes them in terms made “quite clear” – a presumption built on Huakina Development Trust v Waikato Valley Authority [1987] 2 NZLR 188 (HC), a decision Jack Hodder KC describes as for some time “a conspicuous outlier” (at footnote 135). Yet, the decision has since been approved by the Supreme Court in Trans-Tasman Resources Ltd v Taranaki-Whanganui Conservation Board [2021] NZSC 127 with, as Hodder notes, “the legislative foundations of Lands” not mentioned.

On Hodder’s assessment, the Privy Council’s rule in Te Heuheu Tukino v Aotea District Maori Land Board [1941] NZLR 590 (PC) – no enforcement of the Treaty without statutory incorporation – “may have been eroded to near irrelevance” without the Court ever confronting it. The presumption achieves by judicial construction what Parliament was asked, in Sir Geoffrey Palmer’s 1985 White Paper, to enact as supreme law. It declined.

So too with tikanga. Our courts, the Chief Justice says, have “found that tikanga has been and will continue to be recognised in the development of New Zealand’s law where it is relevant to the proceeding.” Of Ellis v R (Continuance) [2022] NZSC 114, the lecture relays the Court’s own reassurance: “this was not a revolutionary moment, but simply an example of the common law in operation.” Part of the decision deserves that description. No statute governed whether a criminal appeal survives the appellant’s death, and in fashioning a continuance test, the Court was settling one of the common law’s unanswered questions. Had the judgment stopped there, the lecture’s account would hold.

But it did not stop there. Tikanga entered the case because the Court put it there. At the hearing, “the Court raised the issue of the relevance of tikanga to the question of continuance” ([33]), and the Court directed what the further submissions were to cover ([156]). The Crown’s acceptance of its relevance, which the Chief Justice’s lecture offers as evidence of orthodoxy, came after that direction. What the Court then declared was unnecessary to the result on its own reckoning: Glazebrook J’s tikanga analysis “leads to the same conclusion as that reached without considering tikanga” ([146]), as O’Regan and Arnold JJ confirm ([315]).

And the unanimity the lecture invokes attaches to the modest proposition at [19], recognition “in cases where it is relevant.” O’Regan and Arnold JJ expressly declined to make “any pronouncements of a general nature about the place of tikanga” ([20]). The change was carried by three judges – Winkelmann CJ, Glazebrook and Williams JJ ([21]). The minority named it for what it was: “the majority overrule the established test for incorporation of custom into the common law, but without saying how it will be replaced,” when “no counsel argued that it should be overruled” (at footnote 279).

The overruled test was not the Court’s to unmake. It gave effect to the statutes that made English law the basis of the courts’ common law jurisdiction – the English Laws Act 1858 and its successor, the Imperial Laws Application Act 1988 – statutes the majority never confronted, as Peter Watts KC has demonstrated in an argument examined earlier in this series. A rule no one asked the Court to abandon was abandoned in a case that did not require it, over a statutory foundation Parliament alone may alter.

On the distinction John Finnis drew in his 2015 Gray’s Inn lecture, between developing the common law and judicial legislation, the Supreme Court’s approach in Ellis was the latter. Dressed in the language of continuity, unanimity and caution, the change was made to sound like the common law’s ordinary growth. It was not. It is the courts overthrowing the statutory foundations of New Zealand’s legal order.

And the fifty years of statutes the lecture catalogues, directing judges to weigh tikanga, cut the other way. Field by field, Parliament has performed the translation function the orthodox account assigns it, conferring on tikanga exactly the legal force it chose, where it chose. A court that generalises past every one of those grants is not filling a silence Parliament left but overriding a calibration Parliament made. The Treaty of Waitangi Act 1975 shows how deliberate that calibration was. Parliament created a tribunal to inquire and to recommend, and left the settlement of claims in political hands.

Ellis changed the rule by which law enters New Zealand’s legal order. The Chief Justice’s account classes that change as the common law’s ordinary operation. That classification becomes a licence: it permits the foundations to be altered without the alteration registering as constitutional change.

The second feature that makes the descriptive reading incomplete is that it is not merely description, but commendation. Legislation, court decisions, and historical and cultural events are named in parallel as three forces by which the constitution is “constantly being reshaped” – and the lecture then says that in this fluidity “lies strength.” A mere description does not commend what it describes. To call the reshaping a strength is to approve it, and approval of the courts’ part in it is not an observation about how an uncodified constitution behaves. It is an account of constitutional legitimacy.

Of course, the qualifications are there: the disclaimer of any power to invalidate legislation in each address, and the definitional care in Auckland. But the parallel itself has never been qualified: court decisions ranked with legislation, not beneath it.

The constitution the courts apply

The orthodox account of the settlement, given its foundational expression by Dicey and defended in our own time by Jeffrey Goldsworthy, and for New Zealand by James Allan, permits courts a good deal. They interpret statutes, choosing between available meanings where the words are genuinely ambiguous and sometimes drawing on the principle of legality, the rule that prefers the reading consistent with fundamental rights, to resolve ambiguity. They apply statutory standards such as “reasonable” or “in the public interest” to facts as cases arise. And they settle the common law’s unanswered questions, case by case, in the areas Parliament has not occupied, as they have always done.

Finnis’s distinction is the most useful here. Developing the common law is backward-looking – the court applies the law that was already in place when the cause of action arose, refining at the edges. Judicial legislation is forward-looking – the court selects a new rule because it considers it preferable. The first is legitimate. The second is not.

The Chief Justice’s account treats the absence of a written constitution as room for the courts to share in shaping the constitutional order: a flexible, uncodified settlement is one the judges help to evolve. The logic runs the other way. On Dicey’s account, judicial legislation is “subordinate legislation, carried on with the assent and subject to the supervision of Parliament.” In a codified system, the constitution itself names who may amend it and how. An uncodified system has no such instrument. There is therefore nothing that could license a second constituent actor alongside the legislature. The absence of a charter does not diffuse the authority to reshape the constitution. It concentrates it, leaving Parliament as the sole repository of constituent power, because there is no higher law to which anyone else could appeal.

What we stand to lose

The orthodox account of New Zealand’s constitutional settlement cannot accommodate the constitutional theory the Chief Justice has set out across these addresses. Her theory is revolutionary.

Not “revolutionary” as a term of abuse. Revolutionary in the constitutional sense: it changes who decides what the law is. That is the sense the late Professor Jock Brookfield gave the word, defining a revolution to include constitutional change “contrary to any limitation or rule of change belonging to that legal order” (Waitangi and Indigenous Rights: Revolution, Law and Legitimation (2nd ed, Auckland University Press, 2006) at 22). It is the definition Peter Watts KC brings to Ellis. The definition requires no power to strike down statutes. The theory rests on a warrant of the office’s own writing: the authority for the wider account of the court’s role is nothing more than the court’s own account of it.

This is not a quibble about the outcomes of individual cases. The objection is to the method the outcomes share. Case by case and speech by speech, it relocates constitutional authority itself – the power to decide who reshapes the constitution and on what warrant – from the body the voters elect to the one body they cannot replace at an election. Accomplished not by legislators voters can turn out, but by judges they cannot.

Three decisions this column has examined show the method at work. No judgment cites the Chief Justice’s addresses. None needs to. Each decision is regular only under the Chief Justice’s wider account of the court’s role. When the court reads “must” to mean “must, unless,” as in Fitzgerald v R; when it draws a mandatory climate duty from a purpose provision, as in Climate Clinic; when it installs a source of law on its own initiative in an appeal that did not require it, as in Ellis – the words on the page cease to be the measure of the law.

On this approach, the lawyer advising on a major project, the Crown counsel briefing the Attorney-General on a Bill, the District Court judge applying a statutory penalty, the citizen reading a statute on the parliamentary website: none of them can take the words on the page as final.

The deeper cost is to democratic accountability. The settlement rests on a single proposition: ultimate authority over constitutional change belongs to the institution the voters can replace. Parliament can be thrown out. The courts cannot.

This is not a concern of one political tradition. Cullen named the remedy: “It is for the public to grant the Courts a larger constitutional mandate; not for the Courts to build one upon an interpretation of constitutional history.”

And Cullen’s second warning has matured alongside the first: advocacy of constitutional change by sitting judges invites the inference of partiality. The inference does not need to be drawn, only to be available. The concern is not actual bias. It is that a sitting Chief Justice has publicly commended a method of constitutional development that will itself be contested in cases before her court. The losing party in the next constitutional case has been given, by the head of the judiciary, reason to wonder whether the outcome was ever open.

An earlier column in this series argued that the methodology which produced Ellis has a father, Lord Cooke of Thorndon, and that the present court is his product. The Chief Justice has now articulated, across three addresses, the theory her court inherited. Goldsworthy anticipated the ending: a successful legal revolution is never acknowledged as one, but depicted as “a judicial discovery, or rediscovery, of what the law had always been.” The braided river is that depiction, offered in advance.

The speeches have been in plain sight. It is an unusual revolution that need not be concealed.