Home
| Databases
| WorldLII
| Search
| Feedback
Otago Law Review |
The Dead Hand of the Bill of Rights?
Is the New Zealand Bill of Rights Act 1990 a Substantive
Legal Constraint on Parliament’s Power to Legislate?
Claudia Geiringer[*]
The genesis of the New Zealand Bill of Rights Act 1990 is well known.[1] In 1985, the newly elected Labour Government produced a White Paper, attaching a draft bill of rights for New Zealand.[2] This bill of rights was to be the “supreme law of New Zealand”, entrenched against amendment by simple majority and capable of rendering ineffective ordinary legislation.[3] The public reaction to the White Paper proposal was, however, one of hostility mingled with indifference.[4]With respect to the former, concern centred on the extent to which a bill of rights would result in a transfer of political power to the judiciary.[5] In that light, the Government re-worked the draft bill of rights to resemble an “ordinary statute”.[6] Clauses entrenching the bill of rights and stipulating its supreme law status were removed.[7] Following select committee consideration, a further clause was added to preclude any lingering possibility that the courts would refuse to apply legislation in reliance on the bill of rights (this became section 4). [8]
During the course of its passage through the House of Representatives, the sponsoring Minister, the Rt Hon Geoffrey Palmer, emphasised that the courts would still have important powers to promote conformity with the rights and freedoms contained in the bill of rights. Ultimately, however, Palmer stressed that this was to be a “parliamentary” bill of rights that would affirm the role of Parliament itself as a guardian of fundamental rights and freedoms. Parliament’s responsibilities in this regard would be advanced by a statutory requirement that the Attorney-General report to the House of Representatives if infringing legislation was introduced to the House (this became section 7), as well as by amendments to Standing Orders to empower a select committee to review all legislation against bill of rights standards.[9]
Palmer ’s “parliamentary” bill of rights was duly enacted as the New Zealand Bill of Rights Act 1990 (hereafter, “the Bill of Rights”).[10] Crucial to any meaningful assessment of its impact and effect is an accurate appreciation of the obligations of Parliament under it. There is, however, some lingering confusion over what these might be. It is generally conceded that the Bill of Rights applies to the work of the House of Representatives and its committees and that, accordingly, the House must conform its deliberative processes to it.[11] But what of the content of legislation itself? To what extent does the Bill of Rights operate as a substantive legal constraint on Parliament’s power to legislate?
On this point, two conflicting accounts appear to have emerged: a weak account and a strong account. The weak account goes something like this. The Bill of Rights is a declaration of the human rights that are held to be important (or perhaps even “fundamental”) in New Zealand society. It contains a set of explicit criteria against which the morality, or propriety, or advisability, or perhaps even “constitutionality” of legislation can be measured. The Attorney-General’s reporting obligation under section 7 serves to promote this standard-setting function. It ensures that, if legislative majorities wish to enact Bill of Rights- inconsistent legislation, they must do so explicitly and must weather the political consequences. Ultimately, however, the constraints on the enactment of infringing legislation are political rather than legal. Parliament has retained its sovereign right to legislate in conscious disregard of the Bill of Rights.
This account underlies the bulk of academic commentary on the role of the Bill of Rights in the legislative process,[12] and accords most closely with the actual behaviour of New Zealand’s elected representatives in the post-1990 legislative environment.[13] It is not, however, the account that is given in either of the two leading textbooks on the Bill of Rights.[14] The authors of these textbooks instead
invoke the language of section 3(a) of the Bill of Rights, which says that the Act applies to “acts done ... by the legislative ... [branch] ... of the government of New Zealand.” They say that the act of legislating is an “act done by the legislative branch” and that accordingly, Parliament is under a legal obligation to legislate consistently with the Bill of Rights (albeit an obligation that the courts cannot enforce by reason of section 4).[15]
The proponents of this alternative, “strong” account of the role of the Bill of Rights in the legislative process do little more than hint at its constitutional implications and practical ramifications. This article concludes that they are significant and that, if the strong account is correct, a sizeable gulf has opened up between legal obligation and actual practice.
The question of which account is correct is, therefore, an important one. In addressing that question, this article notes the considerable attraction of the strong account as a way of giving real bite to the idea of a self-enforced, “parliamentary” bill of rights. Ultimately, however, it concludes that it is doubtful whether the strong account can be sustained. Section 3(a) needs to be interpreted in light of the legislative history of the Bill of Rights and the clear legislative policy of continuing legislative supremacy that underpins it. Interpreted in that light, it is doubtful whether the phrase “acts done by the legislative branch” acts as a substantive constraint on Parliament’s law-making powers. Even if section 3(a) purports to do so, it is equally questionable whether it succeeds. That is because
it is doubtful whether section 3(a) succeeds in overcoming the orthodox constitutional proposition that Parliament cannot bind its successors.
Given this doubt over whether the strong account can be sustained, it is suggested that it may be preferable for theorists to concentrate their efforts on developing an account of the role and impact of the Bill of Rights that accords more closely to the actual practice of legislators in the post-1990 political environment.
The clearest expositions of the strong account are found in Paul Rishworth’s contributions to the textbook he jointly authored in 2003,[16] and in Andrew Butler and Petra Butler ’s more recent textbook.[17] These authors hint at a range of supporting justifications for the strong account, which are explored in Part D below. In essence, though, the claim is a simple one. Section 3(a) of the Bill of Rights says that it applies to “acts done ... by the legislative, executive or judicial branches of the government of New Zealand.”[18] The enactment of legislation is such an act and is done by the legislative branch. Accordingly, legislation “ought not to be inconsistent on its face, nor should it authorise acts by others that would be inconsistent”.[19]
Section 4 of the Bill of Rights does not, the authors say, undercut this analysis. Section 4 says that “no court shall ... hold any provision of [an] enactment to be impliedly repealed or revoked, or to be in any way invalid or ineffective”, nor “decline to apply any provision of an enactment”, by reason only of inconsistency with the Bill of Rights. Section 4 thus prevents the courts from attaching certain consequences to Bill of Rights inconsistency. It does not, however, “excuse the legislature from its primary obligation of acting consistently with the Bill of Rights”.[20] The authors draw a distinction in this respect between the substantive
obligations placed on Parliament by the Bill of Rights and the remedial consequences of default. Section 4, they say, is addressed only to the latter. Parliament can and has bound itself to follow the law and must, therefore, observe the Bill of Rights “of its own motion”.[21] The section 3(a) obligation is “only imperfectly enforced” but is “an obligation nonetheless”.[22]
The essence of the strong account is thus that, by virtue of section 3(a) of the Bill of Rights, Parliament has “bound itself” to refrain from enacting legislation that is contrary to the Bill of Rights and that, if it nevertheless does so, it is in breach of a statutory obligation.
The conceptual distinction on which the strong account rests between the substantive content of a legal obligation and the availability of a judicial remedy is unremarkable.[23] The aspect of the strong account that is rather more constitutionally adventurous is the proposition that the Bill of Rights places legal (albeit unenforceable) limits on Parliament’s law-making power. This aspect of the strong account is surely in tension with section 15 of the Constitution Act 1986 (which says that Parliament continues to have full power to make laws) and with orthodox constitutional doctrine, including the proposition that Parliament exercises unlimited law-making powers and its concomitant, that one Parliament cannot bind the legislative powers of the next.[24] It flows from this doctrine that it is meaningless to speak of Parliament being “bound” to legislate in a manner that is substantively consistent with an earlier statute because the very act of legislating inconsistently with the earlier statute is, in law, self-legitimating. In short, by legislating inconsistently with the earlier statute Parliament does not break the law; it changes the law.[25]
Rishworth and Butler/Butler expressly eschew the language of “supremacy” or “superiority” in exploring the status of the Bill of Rights.[26] However the strong account of the role of section 3(a) in “binding” or “obligating” Parliament to legislate in a manner that is substantively consistent with the Bill of Rights canonly be made sense of as a claim for supremacy of a kind: it can only overcome the constitutional objections just articulated if the Bill of Rights is perceived of as hierarchically superior to other legislation. This hierarchical approach is foreign to orthodox Westminster constitutionalism, which views “constitutional” statutes as no different in formal status to any other kind.[27] Such an approach has, however, gained traction in a recent strand of jurisprudence from the United Kingdom and Canada, as well as in one New Zealand decision.[28]
Thus, in R v Secretary of State for Transport, ex parte Factortame Ltd (No 2) the House of Lords famously held that the effect of legislation stipulating that enactments are to be construed and to have effect subject to the provisions of European Community law is that Community law is “supreme” and will prevail over domestic statutes (including subsequently enacted statutes) in cases of conflict.[29] In Thoburn v Sunderland City Council, Sir John Laws expressly articulated the hierarchical approach on which this jurisprudence rests and suggested that the same approach should apply to other “constitutional” statutes.[30]
A similar approach was taken in R v Drybones to the status of the Canadian Bill of Rights, an “ordinary” statute that purports to require the courts to construe and apply legislation so as not to infringe the rights and freedoms contained in it unless the infringing legislation expressly declares that it is to operate “notwithstanding” the Canadian Bill of Rights.[31] The Canadian Supreme Court held in Drybones that this language has the effect of rendering inoperative subsequent, infringing statutes that do not comply with the “notwithstanding clause” requirement.[32] The Canadian Supreme Court has similarly upheld the efficacy of primacy clauses in provincial human rights statutes,[33] and has said that even without an express primacy clause, human rights legislation will take precedence over subsequent, inconsistent statutes because it is “of a special nature and declares public policy regarding matters of general concern.”[34]
It is within this strand of jurisprudence that Elias CJ’s judgment in R v Pora can be located.[35] In Pora, the Chief Justice posited that section 4(2) of the Criminal Justice Act 1985, which purported to prohibit the retrospective application of a criminal penalty “notwithstanding any other enactment or rule of law to the contrary”, prevailed over subsequent amending legislation that purported to apply a particular penalty retrospectively.[36] Section 4(2) was the “dominant” provision and the subsequent amendment, being “subordinate”, must give way.[37]
The proposition on which the strong account rests – that section 3(a) binds or obligates Parliament to refrain from enacting legislation that is inconsistent with the Bill of Rights – similarly must assume a supremacy or primacy for the Bill of Rights vis-à-vis other enactments. It must assume that the status of the Bill of Rights is such that it overrides, as a matter of substantive law, the general rule that Parliament can legislate inconsistently with an earlier statute. Otherwise the strong account would be meaningless: it would posit that Parliament is “bound” to refrain from enacting, but lawfully entitled to enact, Bill of Rights inconsistent legislation.
At first blush, the proposition on which the strong account rests would also seem to imply that the Bill of Rights is entrenched, in other words, that it is protected from repeal or amendment through the ordinary legislative process. Logically, if section 3(a) of the Bill of Rights binds Parliament to refrain from enacting inconsistent legislation, that must include legislation that purports to repeal or amend the Bill of Rights itself. After all, what could be more inconsistent with the Bill of Rights than its own repeal? The proponents of the strong account do not, however, pursue its logic that far. Both Rishworth and Butler/Butler accept elsewhere in their texts (but without any reference to their account of section 3(a) as a constraint on Parliament) that pragmatic and political constraints aside, Parliament is lawfully entitled to repeal or amend the Bill of Rights.[38] On this account, therefore, the Bill of Rights is supreme but it is not entrenched.
This raises the question of how to distinguish between the effects of supremacy and the effects of entrenchment. Where does “inconsistent legislation” end and “repeal/amendment” begin? According to orthodox Westminster constitutional doctrine, there is no such line. Legislating inconsistently and repealing/amending are, in effect, the same thing because a subsequent, inconsistent statute impliedly repeals pro tanto (or, perhaps, curtails the operation of) the earlier legislation.[39]
In exploring how to draw this line, however, the line of cases just discussed is, again, helpful.[40] In none of the authorities just canvassed was the legislation to which primacy was accorded entrenched. While thus accepting the susceptibility of such legislation to formal repeal or amendment, the courts nevertheless recognised its primacy by displacing the doctrine of implied repeal.[41] In Thoburn,
Sir John Laws articulated the matter thus:[42]
Ordinary statutes may be impliedly repealed. Constitutional statutes may not. For the repeal of a constitutional Act or the abrogation of a fundamental right to be effected by statute, the court would apply this test: is it shown that the legislature’s actual – not imputed, constructive or presumed – intention was to effect the repeal or abrogation? I think the test could only be met by express words in the later statute, or by words so specific that the inference of an actual determination to effect the result contended for was irresistible.
Paul Rishworth has suggested with respect to this line of cases that, by displacing the doctrine of implied repeal, the courts are essentially recognising the imposition of a special “manner and form” restriction on Parliament’s law-making powers: that if Parliament wishes to override constitutional legislation, it must do so explicitly.[43] This reference to “manner and form” taps into one of the most frequently asserted glosses on the orthodox doctrine that Parliament cannot bind its successors: the assertion that although Parliament cannot substantively limit the law-making powers of future Parliaments, it can place valid and enforceable constraints on the process by which future legislation is enacted.[44]
It is to be noted that neither Elias CJ in Pora nor Sir John Laws in Thoburn accepted this characterisation of their judgments.[45] The Chief Justice in Pora preferred to conceive of the exercise as purely one of interpretation. Faced with two, conflicting statutory directives, the court’s job was to ascertain which one Parliament had intended to be “dominant”. That exercise was to be undertaken
in light of the principle of legality (the presumption that fundamental rights cannot be overridden by general or ambiguous words) but was an exercise of interpretation nonetheless.[46] In Thoburn, Sir John Laws posited that as the doctrine of implied repeal was itself a creature of the common law, there was nothing to stop the courts from modifying it to exclude “constitutional” statutes from its application. Like Elias CJ, however, he expressly rejected the suggestion that this would result in the imposition of a manner and form restriction on Parliament’s law-making powers.[47]
It is at this point that the analogy with Thoburn, Pora and their ilk begins to break down. In all of these cases, the purported primacy of the relevant legislation was recognised and enforced by the judges concerned. In contrast, the strong account posits a primacy for the Bill of Rights that is said to operate as a matter of substantive obligation but without a right of recourse to direct judicial enforcement.[48]
While it is perfectly possible to conceive of a constitutional instrument that is supreme but that judges cannot enforce against the legislative branch,[49] the absence of judicial enforcement does create some challenges for the constitutional rationalisation of the strong account. In particular, neither Elias CJ’s interpretation-based explanation nor Sir John Laws’ common law-based explanation of what is at stake when superiority or “dominance” is accorded to an unentrenched statutory provision can assist. The claim here is not that Parliament has intended the courts to resolve legislative conflicts by according dominance to the Bill of Rights – section 4 dictates quite the opposite. Nor is the
claim that the courts, in exercising their traditional, common law role, will refuse to apply the doctrine of implied repeal to constitutional statutes. Rather, proponents of the strong account must (and do) make a constitutionally bolder claim. They must claim that Parliament is “bound” or “obligated” to enact Bill of Rights-consistent legislation. They must thus claim a legal impediment on Parliament’s law-making powers. Whether this account can ultimately be sustained is discussed further in Part D. First, however, the implications of this account for Parliament and other constitutional actors need to be explored.
In light of the remedial constraint effected by section 4 of the Bill of Rights, why does it matter whether the strong account is or is not the right one? What differences, if any, would one expect to see in the practical operation of the Bill of Rights were the strong account to be generally accepted as correct? These questions are not addressed in any detail by its proponents.[50]
It is suggested here that the implications of the strong account are significant. Given that the account speaks primarily to the legislative process, it is in this respect that its implications are the most far-reaching. A number of potential downstream implications for the courts (and for others involved in the interpretation and application of statute law) are, however, also touched on below.
The only provision in the Bill of Rights that speaks directly to the legislative process is section 7. It obliges the Attorney-General, on or shortly after the introduction of a new Bill to the House of Representatives, to bring to the attention of the House any provision in the Bill that “appears to be inconsistent” with the rights and freedoms contained in the Bill of Rights.
In his exposition of the strong account, Rishworth notes (without exploring the proposition) that the Attorney-General’s obligation under section 7 may be understood as “assisting Parliament to discharge its obligations under section 3(a)”.[51] However, the subsequent chapter on the Attorney-General’s reporting function is under different authorship and is more consistent with the weak account of section 3(a).[52] Butler and Butler ’s chapter on section 7 is likewise more consistent with the weak account. On this weak account, the section 7 vetting process acts as a deterrent to the enactment of infringing legislation by fostering the development of a political rights culture, by minimising the risk that Parliament will inadvertently enact infringing legislation and by heightening the political consequences of deliberate, legislative disregard. Ultimately, however, these constraints are institutional and political rather than legal: Parliament is still “free to legislate in a manner that is inconsistent with the Bill of Rights”.[53]
In contrast, the very essence of the strong account of section 3(a) is that it elevates the proposition that legislators ought not to enact Bill of Rights- inconsistent legislation to the status of a legal obligation. This does not mean that Parliament is bound to accept the Attorney-General’s advice in every case. There is always room for disagreement as to the content and application of human
rights protections – even more so, as to the application of the broad reasonableness standard found in section 5 of the Bill of Rights.[54] Further, under section 3(a), it is surely Parliament, not the Attorney-General, that is ultimately and authoritatively empowered to determine whether or not legislation is Bill of Rights-compliant.[55] What members of Parliament must not do, if the strong account is correct, is vote for legislation that appears to them to infringe the Bill
of Rights.
The positive corollary of this negative proposition is surely that members of Parliament must satisfy themselves of the Bill of Rights consistency of all legislation that they intend to support. This has procedural implications. How the House of Representatives organises itself to discharge its commitments under section 3(a) is, first and foremost, the prerogative of the House. At a minimum
however, the House must surely be required to have adequate mechanisms in place to ensure the discharge of those commitments in good faith and in a manner that reflects their obligatory character.
In sum, then, the strong account of section 3(a) has both substantive and procedural implications. Members of Parliament must not vote for legislation that appears to them to infringe the Bill of Rights. It follows that members must satisfy themselves of the Bill of Rights consistency of legislation that they intend to support. It also follows that the House must have adequate mechanisms in place to ensure the discharge of those commitments in good faith. Quite clearly, this is not how the role of the Bill of Rights is currently understood within the precincts of Parliament. In the 2005 edition of Parliamentary Practice
in New Zealand the House’s principal advisor, the Clerk of the House of Representatives, describes the role of the Bill of Rights in the legislative processas follows: [56]
The Bill of Rights Act effectively sets out a “checklist” of most of the important public law tests that could be devised for a proposed law. This does not mean that Parliament will never decide to legislate inconsistently with the Bill of Rights Act. It may, in an appropriate case, decide to do this if it considers that the circumstances warrant it.
The express pronouncements of legislators themselves when justifying a decision to proceed with legislation in the face of a section 7 report likewise reflect the belief that Parliament is ultimately entitled to enact legislation that infringes the Bill of Rights. Admittedly, legislators do sometimes rationalise a decision to proceed with such legislation on the basis that they disagree with the Attorney-General’s advice.[57] This is consistent with the strong account. On other occasions, however, legislators are openly assertive of their perceived entitlement to legislate in contravention of the Bill of Rights.[58] On yet other occasions, legislators are vague as to whether or not they believe they are overriding the Bill of Rights when they enact legislation in the face of a section 7 report. They assert a policy justification for proceeding with the legislation but fail to clarify whether, in light of this justification, they consider the legislative proposal to be reasonable within the terms of section 5 of the Bill of Rights.[59] If legislators regarded Bill of Rights compliance as obligatory, one would not expect to see this degree of equivocation.
Turning to the mechanisms that are in place to facilitate Bill of Rights scrutiny, it is doubtful whether they are sufficient to ensure the good faith discharge of what, on the strong account, are Parliament’s legal obligations. With the exception of Standing Order 266, which facilitates the Attorney-General’s section 7 reporting obligation, the Standing Orders of the House of Representatives make no special provision for the scrutiny of legislation for Bill of Rights consistency.[60] Select Committees are, of course, entitled to consider the Bill of Rights implications of legislation that has been referred to them but are not obliged to do so.[61] Against this background, the Speaker of the House ruled in 1991 that whether legislation is inconsistent with the Bill of Rights is “not a matter of order or responsibility of the House”, a conclusion that is hard to reconcile with the strong account of section 3(a).[62]
In the absence of formal mechanisms for parliamentary scrutiny of legislation for Bill of Rights compliance, the adequacy of such scrutiny in any particular case is dependent on the availability and quality of the advice received by the Attorney-General under section 7, on the nature and quality of any public submissions that are made to select committees, and on the skills and predilections of particular members of Parliament. This is not the place for a thoroughgoing analysis of each of these variables. The general point, however, is that while there is considerable variation, select committee reports and parliamentary debates provide little evidence of systematic and comprehensive parliamentary scrutiny of the Bill of Rights implications of legislation.
While this is a matter of general concern, the concern is most acute where there has been no assessment by the Attorney-General and her advisors of the Bill of Rights implications of legislation, or where Parliament decides to proceed with legislation in the face of a section 7 report. The former situation can arise in relation to amendments to legislation proposed during the course of its passage through the House. These do not attract formal scrutiny under section 7,[63] a loophole that parliament has not seen fit to address despite pleas from various quarters.[64]
As to the latter situation, while legislators are clearly free to disagree with the Attorney-General’s advice, if the strong account is correct one would only expect them to do so after careful, independent consideration of the Bill of Rights issues raised and, preferably, on the basis of an alternative source of credible advice. Anything less surely places a serious question mark over the good faith discharge
of the section 3(a) obligation.
Even on the weak account of the role of the Bill of Rights in the legislative process, there is a serious question as to whether the House’s current posture towards the Bill of Rights represents an adequate response to the moral, constitutional and international imperatives that underlie it. The lack of any formal, parliamentary mechanism for Bill of Rights scrutiny is contrary to the vision of the Labour government that introduced the Bill of Rights,[65] and is at odds with the practice in comparable jurisdictions such as the United Kingdom and Canada.[66] The enactment of Bill of Rights-inconsistent legislation almost inevitably places New Zealand in default of its international treaty obligations and is, increasingly, at odds with common law constitutional values being developed by the courts.[67] The idea that legislators would enact such legislation deliberately or even indifferently ought to be unpalatable to the New Zealand public, particularly given that provision for reasonable limits on rights is built into the structure of the Bill of Rights itself.[68]
The essential difference between the strong and weak account, however, is that whereas on the weak account, such disregard or indifference may amount to a breach of a moral or constitutional imperative, on the strong account, it amounts to a breach of a domestic legal obligation.
If the strong account is correct, the practice of successive executive governments under the Bill of Rights would likewise seem to fall short of legal requirements. Governmental guidelines on the development of legislative proposals operate on the assumption that Bill of Rights-inconsistent legislation is a potential, if somewhat regrettable, outcome.[69] This is inconsistent with the strong account, as is the very fact that successive Cabinets have been prepared to sponsor legislation in the face of advice from the Attorney-General that it infringes the Bill of Rights.[70] Whereas members of Parliament are entitled to disagree with the Attorney- General’s advice, serious rule of law concerns must surely arise from Cabinet disregarding the advice of its own senior law officer on a matter of legal obligation.
The strong account also has implications for the Attorney-General him or herself. If the strong account is correct, the Attorney-General is surely not able to vote for legislation that, on his or her own advice, is contrary to Parliament’s binding legal obligations. As suggested above, the very essence of the strong account is that members of Parliament (the Attorney-General included) are required to refrain from voting for legislation that appears to them to infringe the Bill of Rights. However, given the convention of collective Cabinet responsibility, as long as the Attorney-General is a Minister inside Cabinet, he or she cannot vote against a government-sponsored Bill.[71] Faced with such a predicament, it is suggested that an Attorney-General’s only option is to resign. It is instructive in this regard that in Canada, where no government has ever proceeded with legislation that would attract the equivalent of a section 7 report, a number of commentators have suggested that were a government to do so, the Minister of Justice (who performs the equivalent role to our Attorney-General) would feel compelled to resign.[72]
Judicial references to the role of the Bill of Rights in constraining the enactment of legislation are rare. Of those that do exist, most are consistent with the weak account.[73] Exceptionally, in Pora, the Chief Justice signalled her adherence to the strong account. She said that if Parliament had enacted legislation that infringed the Bill of Rights, it would be “in breach of the obligation recognised by s 3 (although the Court must give effect to the legislation)”.[74]
The dearth of judicial authority on the role of section 3(a) in constraining Parliament’s law-making powers is perhaps unsurprising. The remedial constraint effected by section 4 of the Bill of Rights is such that the implications of the strong account are far less dramatic for the courts than for those involved
directly in the legislative process. Nevertheless, there are a number of ways in which acceptance of the strong account may potentially influence the development of Bill of Rights jurisprudence. For want of space, this article is able to do little more than raise a series of issues that the author considers are deserving of further attention.
Section 6 of the Bill of Rights embodies the principle of legality, that is, the presumption that statutes are capable of a rights-consistent interpretation.[75] It says that wherever an enactment “can” be given a meaning that is consistent with the rights and freedoms contained in the Bill of Rights, that meaning is to be preferred to any other meaning.
Rishworth suggests that the strong account of section 3(a) renders section 6 of the Bill of Rights redundant and he is surely right.[76] It follows from the proposition that Parliament is legally obliged only to enact legislation that is consistent with the Bill of Rights that those charged with the interpretation of legislation ought to presume such consistency.[77] Of course, one might go further and suggest that section 3(a) is also redundant in this respect. A presumption of consistency would likely have evolved in reliance on a bare statement of rights and freedoms, supported by the common law principle of legality.[78]The more interesting question, however, is whether, if the strong account is correct, section 3(a) may have a galvanising effect on the application of section 6 of the Bill of Rights. On the strong account, the courts’ interpretative power under section 6 may be all that stands between Parliament and the dereliction of a legal duty. That conception of the role of section 6 may be thought to support a greater degree of stringency in its application than is evident from current case law.[79] For example, there might be a question as to whether case law limiting the reach of section 6 to the adoption of meanings that the words of a statute can “reasonably” or “properly” bear ought to be discarded,[80] as it has been in United Kingdom.[81]
It will be remembered that section 4 of the Bill of Rights says that no court shall hold any provision of an enactment to be “impliedly repealed or revoked, or to be in any way invalid or ineffective”, nor “decline to apply any provision of an enactment” by reason only of inconsistency with the Bill of Rights. The strong account is premised on a restrictive reading of section 4 that emphasises two features of it.
The first is that section 4 is directed only at the courts and, accordingly, does not relieve Parliament from its obligations under section 3(a). By extension of this logic, section 4 must likewise be irrelevant to the obligations of other persons or bodies to which the Bill of Rights applies by virtue of section 3(a) and (b).[82]These may include government officials, quasi-judicial tribunals, the House of Representatives and its members, officers of Parliament such as the Controller and Auditor-General, and non-governmental bodies that exercise statutory functions.
If these persons and bodies are not bound by section 4, does that mean that some or all of them are entitled and/or required to accord primacy to the Bill of Rights over other enactments and, thus, to refuse to apply legislation that infringes the Bill of Rights? In Canada, the Supreme Court has held that administrative tribunals that have jurisdiction (express or implied) to decide questions of law arising under a legislative provision are presumed also to have jurisdiction to decide the constitutional validity of that provision.[83] Is that also the case in New Zealand? If so, a further issue may arise as to whether, in light of section 4, a court on judicial review of the decision of such an administrative tribunal would be required to uphold or overturn such a finding of unconstitutionality.
The second feature of section 4 that is emphasised by the strong account is that section 4 does no more than to remove certain, specific remedial consequences from the courts’ repertoire. That being so, any remedies for Bill of Rights-inconsistent legislation that are not listed in section 4 are potentially available.
The obvious contender in this respect is the judicial power to grant declarations of legislative inconsistency with the Bill of Rights. It is likely that the New Zealand courts will, in any event, recognise the availability of this remedy. Case law to date has been generally supportive but the position has not been conclusively
established.[84] If the strong account of section 3(a) is correct, the case for its availability is unshakeable – such a remedy is nothing more nor less than a declaration of Parliament’s legal obligations.
Might there be other remedies for legislative inconsistency with the Bill of Rights that have been omitted from the apparently comprehensive list contained in section 4? One possibility that, on the strong account, may require more careful exploration is whether public law compensation ought to be available for legislative inconsistency with the Bill of Rights.[85] Compensation is not, after all, one of the remedies listed in section 4 and its availability might be thought to flow from the proposition that Parliament, in enacting Bill of Rights-inconsistent legislation, is in breach of a legal obligation. Acceptance of the strong account would at least require the merits of this argument to be re-evaluated.[86]
Finally, a third feature of section 4 ought to be highlighted. Section 4 purports to preclude the invalidation of enactments “by reason only” of inconsistency with the Bill of Rights. It therefore leaves open the possibility of judicial invalidation of legislation by reason in part (but not only) of Bill of Rights inconsistency. The strong account of section 3(a) may offer a partial foundation for such judicial invalidation.
This logic may provide a supporting rationale for the result reached by Elias CJ in Pora.[87] As discussed above, the Chief Justice would have disapplied a provision in a subsequent, amending statute in order to preserve the full application of section 4(2) of the Criminal Justice Act 1985. That provision embodied a fundamental human rights protection that was also enshrined in the Bill of Rights: the prohibition against retrospective application of a penalty.[88]The Chief Justice’s approach has attracted criticism from some quarters, both because of a general perception that it subverts parliamentary sovereignty and because of a specific concern that it undermines the policy behind section 4 of the Bill of Rights.[89] If the strong account is correct, however, both these objections would seem to dissolve. On that account, the Bill of Rights has primacy over other enactments. Although, the courts cannot enforce that primacy directly, they may be able to do so indirectly. In Pora, the Chief Justice would have done so by recognising and enforcing the primacy of another provision in an
“ordinary” statute that embodied the protected right.[90] The language of section 4, strictly construed, does not preclude such a result.[91]
Admittedly, Elias CJ does not articulate her decision in this way but it is perhaps no coincidence that she does, in the course of her reasoning, signal her adherence to the strong account.[92] If the strong account is not essential to her argument, it at least provides significant reinforcement.
Finally, it should be noted that cutting across much of the above analysis is the possibility that the strong account might be treated as justifying, not greater judicial activism in Bill of Rights cases, but greater judicial deference. The vision on which the strong account is premised of a “parliamentary” bill of rights over which Parliament wields independent, interpretative authority may, in fact, generate a degree of judicial hesitation in overruling or contradicting legislative judgement in some circumstances. That might particularly be so in relation to legislative judgements of the justifiability of prima facie limits on rights and freedoms under section 5 of the Bill of Rights, which inevitably involve a high policy component. Such judicial deference is, however, unlikely to (and perhaps, ought not to) manifest in the absence of clear evidence of an assumption of responsibility on behalf of legislators for ensuring the consistency of legislation with the Bill of Rights.[93]
It is clear, therefore, that the account that is given of the role of the Bill of Rights in constraining Parliament has important implications for the discharge by each of the three branches of government of their obligations under the Bill of Rights. This is true most particularly for those involved in the legislative process itself but may also be true for the courts and, indeed, for others involved in the interpretation and application of statute law. Which account is correct is, therefore, a question of some interest.
The strong account has considerable attraction as a means of according potency to the rights and freedoms protected by the Bill of Rights while remaining attentive to the primary objection to a fully constitutionalised bill of rights for New Zealand: the accretion of judicial power. It envisages the Bill of Rights as a substantive, legal and constitutional fetter on Parliament’s law-making powers, but as a fetter that must primarily be self-enforced. This gives real bite to Palmer’s notion of a “parliamentary” bill of rights. It also, importantly, brings Parliament’s substantive legal obligations into line with New Zealand’s binding international commitments under the International Covenant of Civil and Political Rights and related international treaties.
Is it, however, sustainable? Proponents of the strong account address this question as an issue of statutory interpretation: does the phrase “acts done by the legislative branch” include legislation?[94] While that question does need to be addressed, there is a broader constitutional question that proponents of the strong account neglect. It is whether Parliament is, in constitutional terms, required to treat section 3(a) of the Bill of Rights as a substantive constraint on its power to enact infringing legislation. These two questions are addressed separately below.
Proponents of the strong account argue that the phrase “acts done by the legislative branch” in section 3(a) of the Bill of Rights must include the act of legislating.[95] In interrogating this proposition, it is helpful to consider, first, the legislative history, secondly, the wider statutory context and finally, the language of section 3(a) itself.
The relevant phrase in section 3 of the Bill of Rights – “acts done ... by the legislative [branch]” – was contained in clause 2 of the draft bill of rights that was attached to the original White Paper.[96] In the context of that original proposal, the language was quite clearly intended to apply to the content of enactments.[97]Indeed, it drew on the comparable provision in the Canadian Charter, which the Canadian Supreme Court had recently affirmed applied to legislation.[98] The wording of the clause was not materially altered when the legislation was subsequently revised to constitute an “ordinary” statute.[99]
In this respect at least, the legislative history of the Bill of Rights favours the strong account. On the other hand, the legislative history also indicates that those responsible for re-drafting the Bill of Rights following the decision not to proceed with a fully “constitutionalised” bill of rights did not intend it to constrain Parliament in its law-making function. In its final report on the White Paper, the Justice and Law Reform Committee recommended the introduction of “a bill of rights which is an ordinary statute, that is, not supreme law and not entrenched.”[100] The Committee in fact appreciated that, in order to achieve this, it would be prudent to amend clause 2 (the clause that was to become section 3).[101] This did not, however, happen. Instead, as introduced to Parliament, the “ordinary law” status of the bill of rights was achieved simply by deleting specific references to its “supremacy”, and by removing the special amendment procedure.[102] Clause 2 remained intact.
It seems that the reason clause 2 was not amended was that officials at the Department of Justice responsible for revising the bill of rights failed to appreciate the argument now being made by proponents of the strong account: that, by virtue of the phrase “acts done by the legislative branch”, the Bill of Rights operates as a constraint on the content of the legislation. In its submission to the Select Committee that considered the Bill, the New Zealand Law Society specifically raised the concern that the reference in clause 2 to the “legislative branch” would constrain the law-making powers of Parliament.[103] The Department of Justice dismissed this concern, responding that there was “nothing
in the Bill to limit or affect the power of Parliament to legislate contrary to the bill”.[104]
On the Select Committee’s recommendation, the concerns of the New Zealand Law Society and others nevertheless led to the addition of a new clause 3A – the clause that would become section 4 of the Bill of Rights. In explaining the reason for this addition, Bill Dillon, the Committee’s Chair, noted that it was intended to “allay the fears still expressed by some that the Bill would be in the nature of overriding legislation rather than legislation in equal standing pari passu with legislation that the House is in the habit of passing regularly”.[105]
Finally, throughout its passage through the House, the Rt Hon Geoffrey Palmer gave an account of the effect of the Bill of Rights largely consistent with the weak account. Thus, for example, in explaining the impact of the section 7 vetting procedure, he noted:[106]
It will ... be for the Parliament to make up its mind and to decide whether it should pass a provision that is contrary to the Bill of Rights. The Parliament still has the power and the capacity to do that, but it will have cause to pause, to reflect, and to make members of the public aware that their rights are being whittled away.
The legislative history is, thus, somewhat equivocal. On the one hand, it is clear from that history that when the phrase “acts done by the legislative branch” was initially included in the legislation, it was understood to cover Parliament’s law-making function. That language was never revised.[107] On the other hand, it is equally clear that neither the officials responsible for revising the Bill of Rights nor the legislators responsible for enacting it in its revised form intended the Bill of Rights to constrain Parliament’s law-making powers. Moreover, section 4 of the Bill of Rights was intended as a patch that would conclusively resolve any remaining argument that the Bill of Rights was substantively or remedially superior to other legislation.
Turning, then, to the statutory context, Butler and Butler make two contextual arguments in favour of the strong account but neither are, with respect, especially convincing. First, they argue that if the weak account were correct, there would be nothing for section 6 of the Bill of Rights to “bite on” (in other words, that the strong account is a necessary prerequisite to the operation of the interpretative presumption).[108] That cannot be right. Section 6 is a freestanding interpretative direction that speaks not to Parliament in its law-making role but to other public actors who have responsibilities for the interpretation and application of statute law.[109] Thus, under the Human Rights Act 1998 (UK), an interpretative direction similar to section 6 of the Bill of Rights operates successfully even though that Act does not bind either House of Parliament.[110]
Secondly, Butler and Butler argue that if the strong account were not correct, section 4 of the Bill of Rights would be redundant.[111] Given the lack of apparent appreciation of the potential effect of section 3(a) at the point in time when section4 was inserted, and given the express statement of the Chairperson of the Select Committee on recommending the insertion of section 4 that it would “allay fears” rather than perform any necessary purpose, this redundancy argument is somewhat unconvincing. Further, if it were to be given any weight, a contrary redundancy argument would need to be overcome. As discussed above, the strong account renders section 6 of the Bill of Rights (the interpretation provision) redundant, arguably suggesting that the weak account is to be preferred.[112] Given the chaotic legislative history of the Bill of Rights, however, it is suggested that redundancy arguments of this kind are not safe.
There is, however, another contextual argument that can be made with respect to section 4 of the Bill of Rights that undercuts the strong account. As seen from the legislative history, section 4 was intended as the embodiment of a wider, legislative policy: that the Bill of Rights was not to be substantively or remedially superior to other legislation. It reflects an underlying constitutional principle – the continuing supremacy of Parliament – and the clear intention of the legislature enacting the Bill of Rights that this constitutional principle was not to be disturbed by the enactment of the Bill of Rights. This legislative policy is quite clearly at odds with the strong account.
Proponents of the strong account point out that the language of section 4 is precisely directed at the courts and at the availability of judicial remedies. It cannot be read as a direct constraint on the substantive obligations of Parliament.[113] That may well be true. Nevertheless, the wider legislative policy that section 4 represents is surely relevant when one comes to construe the language of section 3(a).
Against that background, it is suggested that there ought to be some hesitation in construing section 3(a) so as to operate as a substantive constraint on the content of legislation. While the language of section 3(a) is certainly capable of bearing that meaning (and was intended to do so at the time it was initially inserted into the White Paper) such an interpretation would subvert the clear legislative policy reflected in section 4 of the Act: that the Bill of Rights was not to be substantively or remedially superior to other legislation.
Turning, then, to the arguments made by the proponents of the strong account with respect to the language of section 3(a), Rishworth argues that the reference to the “legislative branch” must be a reference to Parliament as defined in the Constitution Act 1986 and that the only relevant “act” that can be “done” by Parliament is the passage of legislation.[114] It is true that the most logical meaning of the “legislative branch” is Parliament. It is, however, generally accepted (including by Rishworth himself) that the reference to the legislative branch in section 3(a) also captures the various constituent parts of Parliament acting independently (for example, the House of Representatives and its committees).[115]
Accordingly, legislation is by no means the only relevant “act” that might be “done by the legislative branch” for the purposes of section 3(a). By way of comparison, it is not as if the reference to “acts done by the judicial branch” refers to the judicial branch acting in concert. Secondly, both Rishworth and Butler/Butler argue that phrase “acts done” must capture both the process of enactment and the content of legislation itself.[116]
Rishworth acknowledges in this respect that it may seem somewhat strained to regard the content of legislation as an “act” but says that this “pedantic” point is overcome by the interpretative direction in section 6 of the Bill of Rights, which requires a meaning “consistent with the Bill of Rights” to be given to the language of section 3(a).[117] This reliance on section 6 in interpreting the Bill of Rights itself is surely an example of bootstrapping. Nevertheless, there is something in the argument that, as Butler and Butler put it, the phrase “acts done” must capture “outputs” as well as “processes and procedures”.[118] If it did not then by analogy, the substantive outputs of the other branches of government, including (by way
of example) the Standing Orders of the House of Representatives, the common law, and executive outputs such as orders in council, circulars and policies, would all be excluded.
If the general proposition that the substantive outputs of the three branches of government are covered by section 3(a) is accepted, the question becomes, does it necessarily follow that all such substantive outputs must be covered? Butler and Butler are clear on this point. They brook no exception to the proposition that all of the substantive outputs of the three branches of government are covered by section 3(a) of the Bill of Rights.[119] In their view, this includes the entire body of the common law, including in its application between private parties.[120]
Rishworth’s position is, however, more equivocal. In particular, he accepts that section 3(a) of the Bill of Rights does not apply to the common law in its application between private parties. Rather, the Bill of Rights impacts only indirectly on the common law through the common law paradigm of incremental development in light of fundamental values. This model is, he says, preferable, first because it preserves the possibility of incrementalism in developing the common law and secondly, because it is “conceptually inappropriate” to regard the judicial process of determining a case as an “act done by the judicial branch”.[121]
In this author ’s view, Rishworth’s approach to the application of the Bill of Rights to the common law in its application to private parties is to be preferred, not least because it is more consistent with the policy underlying section 3: that the Bill of Rights is to control public rather than private conduct.[122] Once this limit on the scope of section 3(a) is conceded, however, it is surely a short step to accept that the content of legislative enactments may also be excluded from the scope of section 3(a). In light of the legislative policy underlying section 4 and the orthodox constitutional principle that underpins that policy, it is surely equally “conceptually inappropriate” to regard legislation as an “act done” by Parliament for the purposes of section 3(a). Indeed, as discussed above, Rishworth himself acknowledges that the notion that the content of legislation (as opposed to the process of enactment) is an “act done” is at best somewhat strained.
Finally, one last argument must be noted. Rishworth points out that the proposition that for the purposes of section 3(a) of the Bill of Rights, legislation is an “act done by the legislative branch” underpins the operation of provisions in the Human Rights Act 1993, introduced in 2001, which enable complaints of legislative contravention of the right to freedom from discrimination to be made under that Act.[123] This is true and does offer some peripheral support to the strong account. It cannot, however, be determinative of the meaning to be given to section 3(a).
In sum, then, there is certainly an argument to be made that, as a matter of statutory interpretation, section 3(a) of the Bill of Rights does apply to the content of legislation. That argument draws support from the general proposition that section 3(a) covers “outputs” as well as “processes and procedures”, from the assumptions of the drafters at the time the language of section 3(a) was inserted into the White Paper, and from similar assumptions made by the drafters of the Human Rights Amendment Act 2001.
Problematically, however, this reading of section 3(a) is at odds with the vision that underlay the revision and enactment of the Bill of Rights as an “ordinary law” instrument that would not disturb, whether substantively or remedially, continuing legislative supremacy. Interpreting section 3(a) against that background, there is a strong argument to be made that, just as judicial decisions on the content of the common law in its application to private parties are exempted from the scope of section 3(a), so too is the substantive content of legislation enacted by Parliament.
In short, attractive as the idea of a substantively superior bill of rights may seem in theory, that idea is at odds with the legislative policy underlying the creation of the Bill of Rights and is not, ultimately, dictated by the statutory language of section 3(a).
If this view as to the interpretation of section 3(a) is correct there is no need to address the broader constitutional question, which is whether, even if section 3(a) purports to constrain Parliament in its law-making function, Parliament is in fact so constrained? For the sake of completeness, however, it is noted that this result must at least be questionable.
This further issue arises because, as discussed in Part B, in terms of orthodox constitutional doctrine, the enactment of inconsistent legislation by a subsequent Parliament does not amount to a breach of section 3(a) but, rather, to a lawful curtailment of its operation. According to that paradigm, it is meaningless to talk about section 3(a) as “binding” or “obligating” Parliament to refrain from enacting legislation that infringes the Bill of Rights because the very act of enacting such legislation is, in law, self-legitimating. The enactment of Bill of Rights-inconsistent legislation may be morally or perhaps even “constitutionally” regrettable but it is not, in any meaningful sense, a breach of Parliament’s
statutory obligations.
For the strong account to be given meaning, it must therefore be premised on a departure from such orthodox doctrine and on an assertion of the primacy of the Bill of Rights over other legislation as a matter of substantive law. In Part B, this article explored the constitutional basis on which the strong account might thus be said to operate. It suggested that this constitutional claim may receive some degree of support by analogy with the line of cases represented by Factortame, Drybones, and Pora in which the courts have accorded “primacy” to constitutional statutes.
Nevertheless, there remain serious questions over whether the constitutional account given in Part B can ultimately be sustained. First, the gulf between the strong account and the orthodox, constitutional position embodied in section 15 of the Constitution Act 1986 is a wide one. The starting point is that the strong account claims a substantive constraint on Parliament’s law-making power: that Parliament is “bound” not to enact legislation that substantively infringes the Bill of Rights. It is true that the weight of New Zealand authority now supports a “manner and form” exception to the proposition that Parliament cannot bind its successors, but there is no such support for the efficacy of substantive
constraints on Parliament’s law-making powers.[124]
While the strand of jurisprudence discussed in Part B is helpful in teasing out the nature of the claim that is being made by the strong account, ultimately, those cases do not go so far as to claim that Parliament is “bound” to refrain from enacting legislation that is inconsistent with the earlier, “constitutional” statute. Those decisions are generated from the rather different perspective of a judicial authority required to resolve a conflict between two pieces of legislation, each of which it has the power to enforce. In each case, the courts resolved this conflict in light of evolving principles of statutory interpretation and against the, also evolving, background of the common law. Thus in Pora, as discussed above, the Chief Justice was able to claim that the exercise was one of ascertaining,
as a matter of statutory interpretation, which of two, conflicting legislative directives Parliament intended to be “dominant”. In contrast, the strong account must make an explicit claim of a statutory impediment on Parliament’s law- making powers (albeit an impediment that operates at the level of substantive but not remedial law). In that sense, it has no equivalent within New Zealand’s contemporary constitutional jurisprudence.
Secondly, the concession made by the proponents of the strong account that the Bill of Rights can lawfully be repealed or amended is, itself, somewhat problematic. The proponents of the strong account never explain on what basis this concession has been made. If section 3(a) binds Parliament to refrain from enacting Bill of Rights-inconsistent legislation, why do its proponents not also assert that section 3(a) binds Parliament to refrain from repealing or amending the Bill of Rights itself? Presumably, the answer is that section 3(a) must be read subject to underlying constitutional principle, including the proposition that Parliament cannot bind the law-making powers of its successors. If proponents of the strong account accept that this is so, however, on what basis do they claim that section 3(a) nevertheless “binds” or “obligates” Parliament to refrain from enacting infringing legislation?
The only logical answer can be that proponents of the strong account regard this lesser restriction on Parliament’s law-making powers as consistent with underlying constitutional principle because it downgrades the claimed impediment on Parliament’s law-making powers into a procedural or “manner and form” restriction: in order to lawfully enact Bill of Rights-inconsistent legislation, Parliament must do so explicitly. As discussed above, this manner and form analysis was proposed by Rishworth to explain what is at stake in cases such as Thoburn, Pora and their ilk.[125] That explanation perhaps provides some basis for the strong account of section 3(a) to be sustained. It was, however, expressly rejected by the respective justices in Thoburn and Pora.[126] Further, in the case of section 3(a), it is not clear by what mechanism the metamorphosis from a substantive to a “manner and form” restriction can be said to have occurred. The language of section 3(a) itself is, on its face, more consistent with a full-blown, substantive restriction on Parliament’s law-making power and, unlike Thoburn and Pora, the transformation to a “manner and form” restriction does not take place through the medium of judicial interpretation.
Thirdly, given the constitutional novelty of what is proposed, it must surely be relevant that, in enacting the Bill of Rights, Parliament did not intend it to operate as a legal constraint, whether substantive or otherwise, on the power of future Parliaments to enact inconsistent legislation. If such a novel claim of a substantive constraint on Parliament’s law-making powers is to be sustained, one might at least expect that the Parliament purporting to impose the constraint clearly intended to do so.
Finally, in light of this legislative history, it is worth reflecting on the fact that, whatever academic or theoretical conclusion is reached as to the meaning of section 3(a), in practice, it is Parliament’s own view that will be determinative.[127]It is open to the courts to reach a view of the meaning of section 3(a) for the purpose of discharging their own responsibilities under the Bill of Rights. However, legislators themselves have the conclusive power to interpret section 3(a) for the purpose of assessing the impact of the Bill of Rights on the legislative process. As discussed above, legislators and others involved in the political process do not consider themselves bound by the Bill of Rights when enacting legislation. In light of the legislation’s history, there is little prospect of change in this regard, certainly not in the near future. The strong account is, quite simply, not what legislators believe they signed up to.
If legislators are not discharging their legal and constitutional obligations under the Bill of Rights, it is certainly the job of academic commentators to tell them so. If, however, there is real doubt as to the constitutional basis on which the strong account can be sustained, the wisdom of promoting an account that is so markedly at variance with actual experience is perhaps questionable. By doing so, commentators may unwittingly contribute to a culture of complacency, in which the strengths and weaknesses of the particular model of rights protection that New Zealand has adopted are not accurately assessed. In short, the strong account has the potential to be a source of mystification rather than illumination in the important exercise of evaluating the operation and effectiveness of New Zealand’s parliamentary bill of rights model.
This article has tried to take the strong account of the role of the Bill of Rights in the legislative process seriously and to lay bare its constitutional and practical implications. Ultimately, however, no matter how attractive the strong account may seem as a means of energising New Zealand’s somewhat fragile parliamentary bill of rights, it is questionable whether it can be sustained. The account is fundamentally at odds with the policy of continuing legislative supremacy that underlay the enactment of the Bill of Rights and the interpretation of section 3(a) needs to be approached against that background. In that light, it is suggested that, just as judicial decisions on the content of the common law in its application to private parties are exempted from the scope of section 3(a), so too may be the substantive content of legislation enacted by Parliament. Further, even if section 3(a) does purport to bind Parliament in its law-making function, it is not clear that, in constitutional terms, it succeeds in doing so.
Given this uncertainty, and particularly in light of the fact that it is the views of legislators that ultimately determine which account is adopted, the wisdom of promoting the strong account is at very least to be questioned.
[*] Senior Lecturer, Victoria University of Wellington School of Law; (Acting) Co- Director, New Zealand Centre for Public Law. Thank you to Amelia Evans and Grace Rippingdale for their research assistance. For comments on the draft (but without attribution of responsibility) my thanks go to Professor John Dawson, Justice Ellen France, Susy Frankel, Ben Keith, Dean Knight, Justice McGrath, Professor Janet McLean, Dr Nicole Moreham, Dr Matthew Palmer, Steven Price and Professor Michael Taggart.
[1] For more extensive discussion of the genesis of the New Zealand Bill of Rights Act 1990, see Paul Rishworth, “The Birth and Rebirth of the Bill of Rights”, in Grant Huscroft & Paul Rishworth (eds), Rights and Freedoms: The New Zealand Bill of Rights Act 1990 and the Human Rights Act 1993 (Brookers, Wellington, 1995) 1-35; Andrew Butler & Petra Butler, The New Zealand Bill of Rights Act: a commentary (LexisNexis, Wellington, 2005) 11-35; Geoffrey Palmer, New Zealand’s Constitution in Crisis: Reforming our Political System (John McIndoe, Dunedin, 1992) 51-58.
[2] “A Bill of Rights for New Zealand: A White Paper” [1984-1985] I AJHR A.6.
[3] Draft New Zealand Bill of Rights, arts 1 & 28, in Ibid., 10 & 16. For commentary on these articles, see, Ibid., 22 [3.11], 23 [3.15], 53-58 [7.1-7.24] & 68 [10.17-10.18]. On the role of the courts in enforcing this bill of rights, see Draft New Zealand Bill of Rights, art 25, in Ibid., 15; and for commentary, Ibid., 22 [3.10] & 40 [6.2-6.3].
[4] Justice and Law Reform Select Committee, “Inquiry into the White Paper – A Bill of Rights for New Zealand: Interim Report” [1986-1987] X AJHR I.8A.
[5] Ibid.
[6] The language of “ordinary statute” or “ordinary law” appeared in the explanatory note to the New Zealand Bill of Rights Bill 1989 and was used repeatedly by its sponsor, the Rt Hon Geoffrey Palmer, during the Parliamentary Debates: see, eg, (10 October 1989) 502 NZPD 13039; (21 August 1990) 510 NZLD 3760.
[7] New Zealand Bill of Rights Bill 1989. Art 25, relating to the right to apply for a judicial remedy, was also removed.
[8] New Zealand Bill of Rights Bill 1989, cl 3A.
[9] Eg, (10 October 1989) 502 NZPD 13038-13039; (17 July 1990) 509 NZPD 2802; (14 August 1990) 510 NZPD 3449-3450; (21 August 1990) 510 NZPD 3760.
[10] See James Allan, “Turning Clark Kent into Superman: the New Zealand Bill of Rights Act 1990” [2000] OtaLawRw 3; (2000) 9 Otago LR 613, n 2, in which he criticises the use of this abbreviation on the basis that it implies the Bill of Rights is something other than an ordinary statute. I nevertheless prefer the abbreviation partly because of its elegance and partly because it, in my view correctly, locates the New Zealand legislation within a wider international and comparative human rights framework.
[11] Eg, Standing Orders Committee, “Report on the Review of Standing Orders” (1995) XLIII AJHR I.18A 79-80; Philip A Joseph, “Report to the Standing Orders Committee on Natural Justice”, in Ibid., Appendix F, 204 at 214-215; David McGee, Parliamentary Practice in New Zealand, 3rd ed (Dunmore Publishing Ltd, Wellington, 2005) 114; Paul Rishworth, “When the Bill of Rights Applies”, in Paul Rishworth, Grant Huscroft, Scott Optican and Richard Mahoney, The New Zealand Bill of Rights (Oxford University Press, Melbourne, 2003) 70 at 77; Butler and Butler, supra n 1, 89.
[12] Eg, Grant Huscroft, “The Attorney-General’s Reporting Duty”, in Rishworth, Huscroft, Optican and Mahoney, supra n 11, 195 at 195-196; Philip A Joseph, Constitutional and Administrative Law in New Zealand, 2nd ed (Brookers, Wellington, 2001) 1025, 1033 & 1048-1054; McGee, supra n 11, 326; Geoffrey Palmer and Matthew Palmer, Bridled Power: New Zealand’s Constitution and Government, 4th ed (Oxford University Press, Melbourne, 2004) 323.
[13] See below, Part C1.
[14] Butler and Butler, supra n 1; Rishworth, supra n 11.
[15] Butler and Butler, supra n 1, 87-89 [5.3.2 & 5.4.1]; Rishworth, supra n 11, 72.
[16] Rishworth, supra n 11, 72.
[17] Butler and Butler, supra n 1, 87-89 [5.3.2 & 5.4.1]. See, also F M Brookfield,“Constitutional Law” [1990] NZ Recent Law Review 213, 224-225; D M Paccioco, “The New Zealand Bill of Rights Act 1990: Curial Cures for a Debilitated Bill” [1990] NZ Recent Law Review 353, n 33.
[18] Section 3(b) extends the application of the Bill of Rights further to persons or bodies
performing public functions.
[19] Rishworth, supra n 11, 72. See, also, Butler and Butler, supra n 1, 87-89 [5.3.2 & 5.4.1].
[20] Rishworth, supra n 11, 72. See, also, Butler and Butler, supra n 1, 89 [5.4.2].
[21] Butler and Butler, supra n 1, 89 [5.4.2]. Somewhat confusingly, the authors state in the same passage that section 4 of the Bill of Rights “contemplates the passage of BORA-inconsistent legislation”. Read in context, however, these words cannot have been intended to mean that Parliament is legally entitled to enact such legislation. Rather, the authors must mean that section 4 contemplates that on occasion Parliament will act in breach of its legal obligation and that if it does so, the courts’ hands are tied.
[22] Rishworth, supra n 11, 72. See, also, Butler and Butler, supra n 1, 87-89 [5.3.2 & 5.4.1].
[23] The doctrine privileging the proceedings of the House of Representatives from judicial scrutiny provides a ready analogy in a closely related context: see, eg, Bradlaugh v Gossett [1884] UKLawRpKQB 20; (1884) 12 QBD 271; McGee, supra n 11, 631-632. This analogy is adverted to by Butler and Butler, supra n 1, 89 [5.4.2]. See, also, the discussion of duties of imperfect obligation and of constitutional duties in A J Harding, Public Duties and Public Law (Clarendon Press, Oxford, 1989) 25-28 and 74-77.
[24] See, eg, A V Dicey, Introduction to the Study of the Law of the Constitution, 10th ed (MacMillan Education, Basingstoke, 1959) 39-40 & 64-68; H W R Wade, “The Basis of Legal Sovereignty” [1955] CLJ 172, 174; O Hood Phillips, Constitutional and Administrative Law, 7th ed (Sweet & Maxwell, London, 1987), esp, chapters 3 and 4.
[25] Ellen Street Estates Ltd v Minister of Health [1934] 1 KB 590, 597 (per Maugham LJ); Wade, supra n 24, 174-176; Anthony Bradley, “The Sovereignty of Parliament – Form or Substance?” in Jeffrey Jowell and Dawn Oliver (eds), The Changing Constitution, 5th ed (Oxford University Press, Oxford, 2004) 26, 30.
[26] Paul Rishworth, “The New Zealand Bill of Rights” in Rishworth, Huscroft, Optican and Mahoney, supra n 11, 1, 2; Butler and Butler, supra n 1, 5 [1.3.3].
[27] Shaw v Commissioner of Inland Revenue [1999] 3 NZLR 154, 157 [14]; Joseph supra n 12, 483 [14.4]. See, though, R v Fineberg [1968] NZLR 119 in which Moller J accepted that unless or until expressly repealed, section 53 of the New Zealand Constitution Act 1852 placed substantive constraints on the law-making powers of Parliament. Moller J did not, however, address the possibility that subsequent, inconsistent legislation might impliedly repeal section 53: see R v Fineberg (No 2) [1968] NZLR 443, 449 (per Turner J); and B V Harris, “The Law-making Powers of the New Zealand General Assembly: Time to Think About Change” [1984] OtaLawRw 6; (1984) 5 Otago LR 565, 567-569.
[28] See, below n 29-37; Joseph, supra n 12, 461-462 [14.2], referring to the trend towards a hierarchy of legal values.
[29] (1990) UKHL 13; [1991] 1 AC 603, esp 658-659 (per Lord Bridge of Harwich), applying the European Communities Act 1972, s 2(4). See, Bradley, supra n 25, 43-49.
[30] [2002] EWHC 195; [2003] QB 151, 186 [62]. For a fuller discussion of this case, see Rebecca Prebble, “Constitutional Statutes and Implied Repeal: The Thoburn Decision and the Consequences for New Zealand” (2005) 36 VUWLR 291.
[31] R v Drybones [1970] SCR 282. The Canadian Bill of Rights is not to be confused with the subsequently enacted Canadian Charter of Rights and Freedoms.
[32] However, by virtue of the “notwithstanding clause”, this is arguably an example of an express “manner and form” constraint rather than a substantive constraint on legislation: see, Peter W Hogg, Constitutional Law of Canada, 4th ed (Carswell, Scarborough, 1997) [32.3(c) and (d)].
[33] Ford v Quebec [1988] 2 SCR 712; Devine v Quebec [1988] 2 SCR 790. See, Hogg, supra n 32, [12.3(b)].
[34] Winnipeg School Division No 1 v Craton [1985] 2 SCR 150, 156. It should be acknowledged that the leading constitutional scholar, Peter Hogg, regards the reasoning in this decision as “dubious”: Hogg, supra n 32, [12.3(b), n 51]. See, also Insurance Corporation of British Columbia v Heerspink [1982] 2 SCR 145, 157-158 (per Lamer J); Scowby v Glendinning [1986] 2 SCR 226, 236 (per Estey J).
[35] R v Pora [2000] NZCA 403; [2001] 2 NZLR 37, 40-52. The Chief Justice’s approach was supported byt wo of her colleagues and opposed by three, with the seventh justice refusing to express a view on it.
[36] Ibid. Of the seven members of the Court of Appeal, three supported this approach, three rejected it and the seventh preferred to express no view on it.
[37] Ibid, 50.
[38] Paul Rishworth, supra n 26, 2 & 4; Butler and Butler, supra n 1, 9 [1.4.5].
[39] Ellen Street Estates Ltd, supra n 25, 597 (per Maugham LJ); Wade, supra n 24 174- 176; Bradley, supra n 25, 30. For authorities suggesting that if the inconsistency is only partial (as will almost inevitably be the case in the Bill of Rights context) it is preferable to characterise the effect of the subsequent statute as a curtailment of the operation of (rather than implied repeal of) the earlier provision: see, eg, Goodwin v Phillips [1908] HCA 55; (1908) 7 CLR 1, 7; R v Pora, supra n 35, 47 [36]; S G G Edgar, Craies on Statute Law, 7th ed, (Sweet & Maxwell, London, 1971) 374-375.
[40] See the cases discussed at supra n 29-35.
[41] See, eg, Thoburn, supra n 30, 186-187 [63]; Drybones, supra n 31, 294-295; WinnipegSchool Division No 1, supra n 34, 156; Pora, supra, n 35.
[42] Supra n 30, 186-187 [63].
[43] Paul Rishworth, “Interpreting and Invalidating Enactments Under a Bill of Rights: Three Inquiries in Comparative Perspective”, in Rick Bigwood (ed), The Statute: Making and Meaning (LexisNexis, Wellington, 2004) 251, 258.
[44] For evidence that the weight of New Zealand authority now favours this gloss,see, eg, Shaw, supra n 27, 157 [13]; Carter v Police [2003] NZAR 315, 325; Westco Lagan Ltd v Attorney-General [2000] NZHC 1350; [2001] 1 NZLR 40, 61-62 [90]-[92], affirmed in Wishart v Police (27 March 2002) HC, Auckland, A185/01 and Reid v Minister of Labour [2005] NZAR 125, 131; F M Brookfield, “Parliamentary Supremacy and Constitutional Entrenchment: A Jurisprudential Approach” (1984) 5 Otago LR 602; Joseph, supra n 12, 513-547; David McGee, “The Legislative Process and the Courts”, in Philip A Joseph (ed) Essays on the Constitution (Brookers, Wellington, 1995) 84, 85; Palmer and Palmer, supra n 12, 156; Rishworth, supra n 11, 73; The White Paper, supra n 2, [7.10-7.15]. For recent affirmation of this doctrine by the House of Lords, see R (Jackson and others) v Attorney-General [2005] UKHL 56; [2006] 1 AC 262.
[45] Supra n 35, 50 [52]; supra n 30, 184 [59].
[46] Supra n 35, esp 50 [53]. Elias CJ sourced this principle of legality both in the common law and in section 6 of the New Zealand Bill of Rights Act 1990.
[47] Supra n 30, 184-187 [59]-[64].
[48] The question of whether indirect remedies are available is discussed below at Part C3.
[49] Arguably, the Constitution of the United States started out life as just such an instrument: see Larry D Kramer, The People Themselves: Popular Constitutionalism and Judicial Review (Oxford University Press, Oxford, 2004), esp, chapter 3.
[50] See, Rishworth, supra n 11, 72; Butler and Butler, supra n 1, 87-89 [5.3.2 & 5.4.1].
[51] Rishworth, supra n 11, 72.
[52] Huscroft, supra n 12, 195. See, though, Paul Rishworth, “The New Zealand Bill of Rights and the Legislature: Praxis and Pitfalls”, submitted for the Centre for Comparative Constitutional Studies’ conference, “Parliamentary Protection of Human Rights”, held on 20-22 July 2006, paper available as “draft” at http:// cccs.law.unimelb.edu.au/index.cfm?objectId=9E065F87-1422-207C-BAC23913D7DB519D BAC23913D7DB519D (last accessed 11/10/06). In this, more recent conference paper, Rishworth develops an account of the role of the Bill of Rights in the legislative process that is somewhat more, but not entirely, in accord with his strong account of section 3(a).
[53] Huscroft, supra n 12, 195; Butler and Butler, supra n 1, 198 [8.4.1].
[54] Section 5 says that the rights and freedoms contained in the Bill of Rights may be subject to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society.
[55] See Rishworth, supra n 52.
[56] McGee, supra n 11, 326. See, also, Ibid, 328.
[57] Eg, Parole (Extended Supervision) and Sentencing Amendment Bill, as reported from the Justice and Electoral Committee: commentary (presented 14/6/04) 4-5.
[58] Eg, Electoral Amendment Bill (No 2), as reported by the Justice and Electoral Committee: commentary (presented 28/9/2001) 8. See, also, Prostitution Reform Act 2003, ss 12 and 13, expressly empowering territorial authorities to make bylaws that are “inconsistent with the [Bill of Rights]”.
[59] Eg, Land Transport (Street and Illegal Drag Racing) Amendment Bill, as reported from the Law and Order Committee: commentary (presented 29/11/02) 10-11; Criminal Procedure Bill, as reported from the Law and Order Committee: commentary (presented 29/07/05) 11.
[60] Standing Orders of the House of Representatives (2005), SO 266. Compare SO 267, requiring special voting procedures in relation to proposals for legislative entrenchment.
[61] In its most recent review, the Standing Orders Committee rejected a number of recommendations from the Clerk of the House that would have improved this situation: see, Clerk of the House of Representatives, “Review of Standing Orders: Submission”, May 2003, Recommendations 18, 48 and 49, pp 36-38 and 62-63; Report of the Standing Orders Committee, “Review of Standing Orders” (2003) AJHR I. 18B 28 and 51; Standing Orders, supra n 60, SO 190(3). See, though, Standing Orders Committee, supra n 61, 29 and Standing Orders, supra n 60, SO 189, giving effect to the Clerk’s recommendation that “human rights” be added to the express subject areas of the Justice and Electoral Committee: Clerk of the House of Representatives, supra n 61, Recommendation 19, 38-39.
[62] (3 July 1991) 516 NZPD 2968. This is the only Speaker ’s ruling on the Bill of Rights.
[63] See, also, Standing Orders, supra n 60, SO 266. Even when the Attorney-General chooses to seek the advice of officials on such post-introduction amendments, there is no routine practice of disclosure of that advice, as there now is for advice received under section 7: see, http://www.justice.govt.nz/bill-of-rights/index.html (last accessed 12/10/06).
[64] See, Standing Orders Committee, supra n 61, 51; R v Poumako [2000] NZCA 69; [2000] 2 NZLR 695, 709 [66] (Henry J) and 717-719 [96] & [101] (Thomas J); Clerk of the House of Representatives, supra n 61, Recommendation 40, pp 57-58; Butler and Butler, supra n 1, 206-207 [8.14.2]. See, though, Hon David Parker, “Speech at the Opening of the Bill Of Rights Act Symposium”, 10/02/06, http://www.beehive.govt.nz/ ViewDocument.aspx?DocumentID=24872, (last accessed 21/11/06), suggesting that there may be some will within Government to improve this situation.
[65] See, supra n 9 and accompanying text; Justice and Law Reform Select Committee, “Inquiry into the White Paper – A Bill of Rights for New Zealand: Final Report” [1987-1990] XVII AJHR I.8C, 11.
[66] See, eg, Janet Hiebert, “Rights-Vetting in New Zealand and Canada: Similar Idea, Different Outcomes” (2005) 3 NZJPIL 63, 75; Antony Lester, “Parliamentary Scrutiny of Legislation Under the Human Rights Act 1998” (2002) 33 VUWLR 1.
[67] Eg, R v Secretary of State for the Home Department, ex parte Simms [1999] UKHL 33; [2000] 2 AC 115.
[68] Bill of Rights, s 5.
[69] Cabinet Office, Cabinet Manual (2001) [5.35]-[5.39]; Cabinet Office, Step by Step Guide: Cabinet and Cabinet Committee Processes (2001), esp [3.58]; Legislation Advisory Committee, Guidelines on the Process and Content of Legislation (2001), chapter 4; and, esp, Ministry of Justice, Guidelines on the New Zealand Bill of Rights Act 1990: A Guide to the Rights and Freedoms in the Bill of Rights Act for the Public Sector (2004), esp Part II, advising policy makers who are “developing or amending legislation that intends to override the Bill of Rights” that they must do so “clearly and expressly.”
[70] There have now been eighteen occasions on which governments have introduced legislation to the House that their Attorney-General believed to infringe the Bill of Rights.
[71] Cabinet Manual, supra n 69, 3.21. See, though, Ibid, 3.23-3.24, the “agree to disagree” provisions, which may give rise to a limited exception if the Attorney-General is appointed from a minor coalition partner. See, also, Palmer and Palmer, supra n 12, 85-89.
[72] Hiebert, supra n 71-72 & 90; Elmer A Driedger, “The Meaning and Effect of the Canadian Bill of Rights: A Draftsman’s Viewpoint” (1977) 9 Ottawa L Rev 303, 311. The Canadian legislature is clearly bound by the Charter in the exercise of its law-making function: see, Constitution Act 1982 (Canada), ss 32(1) & 52(1); Retail, Wholesale and Department Store Union v Dolphin Delivery Ltd [1986] 2 SCR 573, 598- 599 [40]-[41] (McIntyre J).
[73] Eg, Simpson v Attorney-General [Baigent’s Case] [1994] NZCA 287; [1994] 3 NZLR 667, 706 (Gault J); Westco Lagan, supra n 44, 63 [95] (McGechan J); Mangawaro Enterprises v Attorney- General [1994] NZHC 2115; [1994] 2 NZLR 451, 457 (Gallen J). See, also Shaw, supra n 27, 157 [14].
[74] Supra, n 35, 47 [35].
[75] See, eg, Pora, supra n 35, 50 [53] (Elias CJ); Ngati Apa Ki Te Waipounamu Trust v R [2000] NZCA 45; [2000] 2 NZLR 659, 675 [82]; Paul Rishworth, “Interpreting Enactments: Sections 4, 5 and 6”, in Rishworth, Huscroft, Optican and Mahoney, supra n 11, 116, 132-133.
[76] Rishworth, supra n 11, 72.
[77] See, eg, Hogg, supra n 32, [35.5], explaining how a presumption of this kind operates in the context of a fully constitutionalised charter of rights.
[78] See, Rishworth, supra n 75, 132-133.
[79] There is some hint of this in the judgment of Elias CJ in Pora, supra n 35, 47 [35].
[80] Eg, Ministry of Transport v Noort [1992] 2 NZLR 260, 272 (Cooke P); Quilter v Attorney- General [1997] NZCA 207; [1998] 1 NZLR 523, 581 (Tipping J); Moonen v Film and Literature Board of Review [1999] NZCA 329; [2000] 2 NZLR 9, 16.
[81] Eg, Ghaidan v Mendoza [2004] UKHL 30; [2004] 2 AC 557. See, Claudia Geiringer, “It’s Interpretation, Jim, but not as we know it: Ghaidan v Mendoza, the House of Lords and Rights- Consistent Interpretation” in Paul Morris and Helen Greatrex (eds) Human Rights Research: Victoria University of Wellington (Victoria Human Rights Programme, 2005). Note, though, that the Human Rights Act 1998 (UK) does not apply to Parliament.
[82] Section 3(a) applies the Bill of Rights to acts done by the legislative, executive and judicial branches of the government of New Zealand. Section 3(b) applies it to acts done by any person or body in the performance of any public function, power, or duty conferred or imposed on that person or body by or pursuant to law.
[83] Nova Scotia (Workers’ Compensation Board) v Martin; Nova Scotia (Workers’ Compensation Board v Laseur [2003] 2 SCR 504; Paul v British Columbia (Forest Appeals Commission) [2003] 2 SCR 585. See John M Evans, “Principle and Pragmatism: Administrative Agencies’ Jurisdiction over Constitutional Issues”, in Grant Huscroft and Michael Taggart (eds), Inside and Outside Canadian Administrative Law: Essays in Honour of David Mullan (University of Toronto Press, Toronto, 2006)377, 380.
[84] Eg, Moonen, supra n 80, 17 [19]-[20]; Poumako, supra n 64, 715-720 [86]-[106] (Thomas J); Belcher v Chief Executive of the Department of Corrections CA184/05, 19 September 2006, [57]-[59]. See, Butler and Butler, supra n 1, 1026 [28.6.13].
[85] For the, admittedly limited, circumstances in which the Canadian courts are prepared to give damages for action taken under the shield of unconstitutional legislation, see, Mackin v New Brunswick [2002] 1 SCR 405.
[86] See, though, Grant Huscroft, “Civil Remedies for Breach of the Bill of Rights”, in Rishworth, Huscroft, Optican and Mahoney, supra n 11, 811, 818-819 for the view that the award of compensation for a statutorily protected breach of the Bill of Rights would contravene section 4.
[87] Supra n 35.
[88] Bill of Rights, s 25(g).
[89] See, eg, Andrew Butler, “Implied Repeal, Parliamentary Sovereignty and Human Rights in New Zealand” [2001] PL 586; Anita Killeen, Richard Ekins and John Ip, “Undermining the Grundnorm?” [2001] NZLJ 299. Contrast, Daniel Kalderimis, “R v Pora” [2001] NZLJ 369. Note that the Chief Justice’s approach was supported by two of her colleagues and opposed by three, with the seventh justice refusing to express a view on it.
[90] Contrast, Mangawaro Enterprises, supra n 73, 457. See, also, Brookfield, supra n 17, 225.
[91] This argument is acknowledged although not accepted by Andrew Butler in Butler, supra n 89, n 30.
[92] Supra n 35, 47 [35].
[93] Compare R v D [2003] 1 NZLR 41, 47 [29]-[31] with Poumako, supra n 64, 709 [66] (Henry J) and 717-719 [96] & [101] (Thomas J).
[94] Rishworth, supra n 11, 72; Butler and Butler, supra n 1, 87-89 [5.3.2 & 5.4.1].
[95] Rishworth, supra n 11, 72; Butler and Butler, supra n 1, 87-89 [5.3.2 & 5.4.1].
[96] Draft New Zealand Bill of Rights, art 2, in White Paper, supra n 2, 11.
[97] White Paper, supra n 2, 69-70 [10.19-10.21]. See Butler and Butler, supra n 1, 88-89 [5.4.1].
[98] Dolphin Delivery, supra n 72, 598-599 (McIntyre J). See Rishworth, supra n 11, 72, n 13. Note, though, that the comparable provision in the Canadian Charter does not contain the phrase “acts done by”.
[99] The only alteration was the addition of the word “only” to the chapeau clause of section 3: “This Bill of Rights applies only to acts done –”.
[100] Justice and Law Reform Committee, supra n 65, 3. See, also, New Zealand Bill of Rights Bill (as introduced), explanatory note, i.
[101] Justice and Law Reform Committee, supra n 65, 5.
[102] Draft New Zealand Bill of Rights, arts 1 & 28, in White Paper, supra n 2, 10 & 16; New Zealand Bill of Rights Bill (as introduced).
[103] Department of Justice, “Report to Justice and Law Reform Select Committee on the New Zealand Bill of Rights Bill: Part One”, 15 May 1990, 10 & 12.
[104] Ibid.
[105] Bill Dillon, (17 July 1990) 509 NZPD 2799-2800.
[106] Rt Hon Geoffrey Palmer, (21 August 1990) NZPD 510 3760. See, also, Rt Hon Geoffrey Palmer, (20 October 1988) 493 NZPD 7550; Rt Hon Geoffrey Palmer, (10 October 1989) 502 NZPD 13039.
[107] See, Rishworth, supra 11, 72, n 13; Butler and Butler, supra n 1, 88-89 [5.4.1].
[108] Butler and Butler, supra n 1, 87 [5.3.2].
[109] In terms of section 3 of the Bill of Rights, this may include both members of the legislative, executive and judicial branches but also any person or body in the performance of any public function, power, or duty conferred or imposed on that person or body by or pursuant to law.
[110] Human Rights Act 1998 (UK), ss 3 and 6.
[111] Butler and Butler, supra n 1, 87 [5.3.2].
[112] Part C3.
[113] Rishworth, supra n 11, 72. See, also, Butler and Butler, supra n 1, 87-89 [5.3.2 & 5.4.1].
[114] Rishworth, supra n 11, 72.
[115] See supra n 11.
[116] Rishworth, supra n 11, 72, n 13; Butler and Butler, supra n 1, 87-88 [5.3.2].
[117] Rishworth, supra n 11, 72, n 13.
[118] Butler and Butler, supra n 1, 87-88 [5.3.2].
[119] Ibid, 87-96 [5.3]-[5.6].
[120] Ibid, 95-96 [5.6.6]-[5.6.10] and 105-106 [5.8.11]-[5.8.15].
[121] Rishworth, supra n 11, 99-102.
[122] See, White Paper, supra n 2, 69-71 [10.19]-[10.23].
[123] Rishworth, supra n 11, 72, n 13. See, esp, Human Rights Act 1993, ss 20K(1) & 20L(1).
[124] See, supra n 44.
[125] Rishworth, supra n 43.
[126] Supra, n 45.
[127] See, in this respect, Matthew S R Palmer, “What is New Zealand’s constitution and who interprets it? Constitutional realism and the importance of public office-holders” (2006) 17 PLR 1, and the theory of constitutional realism developed there.
NZLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.nzlii.org/nz/journals/OtaLawRw/2007/5.html