REASSESSING THE NEW COMMONWEALTH MODEL OF CONSTITUTIONALISM
Stephen Gardbaum*
(International Journal of Constitutional Law, Vol. 8, No. 2, 2010)
INTRODUCTION
Nearly nine years ago, shortly after the United Kingdom’s Human Rights Act
(HRA) came into effect, I published what I believe was one of the first articles to
identify and present the case for a new, intermediate model of constitutionalism that
in somewhat different versions had been adopted in Canada, New Zealand and the
United Kingdom as an alternative to their traditional principle of parliamentary
sovereignty.1
I termed this experiment “the new Commonwealth model of
constitutionalism.”2
Although as an attempted contribution to general constitutional theory, the
case that I presented for the new model over the two traditional options of judicial
and legislative supremacy was primarily a normative one, parts of it depended on or
assumed certain things about how the model would or did operate in practice.
Specifically, I argued that the new model is normatively appealing to the extent it
effectively protects rights while reallocating power between courts and the political
institutions in a way that brings them into greater balance than under the two more
*
MacArthur Foundation Professor of International Justice & Human Rights, UCLA School of
Law. Earlier versions of this article were presented as my keynote address to the 2009 Protecting
Human Rights Conference in Sydney, Australia, to the Regulatory Networks Faculty at ANU and the
University of Melbourne Law School’s Centre for Comparative Constitutional Studies and at the
University of Toronto Law Faculty’s Constitutional Roundtable, the Cardozo/NYU I.CON
Colloquium on Global and Public Law Theory and the University of Chicago Law School’s
Constitutional Law Workshop. Thanks to organizers of and participants at those events, especially
Hilary Charlesworth, Mattias Kumm, Michel Rosenfeld, Ed Santow, Adrienne Stone, David Strauss,
Joseph Weiler, Lorraine Weinrib, and also to Bruce Ackerman, Rosalind Dixon, Janet Hiebert, Aziz
Huq, Kent Roach and Mark Tushnet, for extremely helpful discussion and comments.
1
Stephen Gardbaum, The New Commonwealth Model of Constitutionalism, 49 AM. J. COMP. L.
707 (2001).
2
Others have subsequently used different terms. Mark Tushnet prefers “weak-form” judicial
review, see Mark Tushnet, Alternative Forms of Judicial Review, 101 MICH. L. REV. (2003); Janet
Hiebert, the “parliamentary bill of rights model,” see Janet Hiebert, Parliamentary Bills of Rights: An
Alternative Model?, 69 MOD. L. REV. 7 (2006); and “the dialogue model” has also gained currency.
For clarification, in using my term I did not intend to suggest any conceptual connection between the
new model and Westminster-based parliamentary systems, so that the model is only open to them, any
more than I intended such a connection between the contrasting “American model” (which I argued
was adopted by most post-1945 constitutions around the world) and countries in the western
hemisphere. The labels were selected for the historical facts of where the models were first – and, in
the case of the new Commonwealth model, still only -- employed. Although only very lightly wedded
to my term, I explain in the course of the article why I think the second term is under- and the third
over-inclusive.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 2
lopsided traditional models. In this sense, therefore, some of the benefits that I
claimed for the new model were potential or theoretical ones. As a result, I always
planned to return to the topic for an assessment of whether practice was living up to
theory and, indeed, this is where much of the expressed skepticism towards the new
model has been focused. But I resisted doing so until there was sufficient basis to
make the project meaningfully based on the evidence. In light of almost a decade of
both additional experience under the new model and sustained academic attention to
it in each country, as well as its subsequent expansion into Australia at the territorial
and state levels3
(and possibly also soon at the national), I think that point has now
arrived. Accordingly, my main goal in revisiting the new model in this article is to
try to evaluate its success and distinctiveness in practice.
The article is organized as follows. The first three sections address
preliminaries and set up the framework for the evaluation in the fourth and fifth. In
Part I, I explain what is new and distinctive about the new model. Here, I also
outline the spectrum of possible and actual institutional variations on a theme that it
permits. In Part II, I briefly refine and restate what (still) seems to me the basic case
for the new Commonwealth model of constitutionalism as an intermediate form in
between the two more traditional ones. This basic case also suggests both how we
would ideally want the new model to operate (also in Part II) and the criteria for
evaluating the success or failure in practice of the versions we actually have, which I
set out in Part III. Parts IV and V form the heart of the article. In Part IV, I apply
these ideals and criteria to assess existing experience, as well as the academic
commentary on it, in Canada, New Zealand, and the UK in order to consider whether
on balance the experiment has been a success and which of the versions is working
the best. Finally, in Part V, I identify what strikes me as the major practical dilemma
that has arisen for the model as a whole and how it might best be resolved. As I
explain in this section, one horn of this dilemma is the result of remedial concerns on
the part of courts where they do not have the power to disapply rights incompatible
statutes. The solution I propose is the presumption of a legislative remedy whenever
the legislature agrees with the courts and amends or repeals the relevant statute. I
also discuss a few other possible amendments or reforms, including some that have
been employed in the two more recent Australian versions, which might further
improve the transition from theory to practice.
As this is a scholarly article and not a detective story, I hope readers will not
be disappointed that “[i]n my beginning is my end.” My general conclusion is that
thus far in its short history, the new model has mostly been moderately successful
and distinctive in practice -- at least in its New Zealand and UK versions – and that,
if the experiment is permitted to continue, a few working adjustments can make it
more so in “time future.” Accordingly, as a third form of constitutionalism, the new
model should be given its rightful place on the menu of options alongside judicial
and legislative supremacy.
3
These are the ACT Human Rights Act of 2004 and the Victorian Charter of Human Rights and
Responsibilities of 2006.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 3
I. WHAT IS NEW ABOUT THE NEW MODEL?
For me, the essential and defining features of the new Commonwealth model
as a general, innovative, and distinctive model of constitutionalism are threefold: (1)
a legalized bill or charter of rights; (2) some form of enhanced judicial power to
enforce these rights by assessing legislation (as well as other governmental acts) for
consistency with them that goes beyond traditional presumptions and ordinary modes
of statutory interpretation; and (3) most distinctively, notwithstanding this judicial
role, a formal legislative power to have the final word on what the law of the land is
by ordinary majority vote. In combination, the first two features distinguish the new
model from the traditional conception of parliamentary sovereignty; the third one
from judicial supremacy.4
The new model is thus intermediate between these two
previously exhaustive polar options. By authorizing some form of constitutional
review of legislation by the courts in the name of rights but by giving the legal power
of the final word to the legislature, its great novelty is to decouple the power of
judicial review from judicial supremacy or finality. I just note here what I will
expand on in the next section. This working definition does not mean that the new
model is all or even primarily about the relationship of courts and legislatures; to the
contrary, it is also about rights consciousness among all three branches of
government and the citizenry.
These defining characteristics are quite general and permit a range of
different particular instantiations, particularly with respect to the second and third
features, some of which have in fact been adopted in various countries. Although the
new model has become increasingly synonymous with “parliamentary bills of
rights,”5
they are, as far as I’m concerned, an important sub-set of it as it is perfectly
possible for all three features to apply to a constitutional bill of rights. Accordingly,
in my original article, I firmly identified Canada along with New Zealand and the
United Kingdom as a member of the new model, indeed the founding member. And
despite a significant – though hardly uncontested -- strand of thought among
commentators on the Charter that Canada effectively now has a constitutional system
4
Very briefly, the model of judicial supremacy posits (1) the constitution, including a bill of
rights, as the supreme law of the land, (2) entrenchment of the constitution and its bill of rights against
ordinary majoritarian amendment, and (3) judicial enforcement via the power to invalidate or disapply
legislation and other inconsistent governmental actions, against whose decisions the political branches
are powerless to act by ordinary majority vote. This third feature is what Mark Tushnet refers to as
“strong-form” judicial review, see Tushnet supra note 2. Although to be sure among the many
countries to have adopted this model since 1945, the specific institutionalization of such judicial
review varies – most notably between centralized and decentralized judicial review – the acceptance
of these three fundamental principles does not; rather these are variations on a theme.
Accordingly, a constitutional bill of rights without a legislative override power that may be
amended by a legislative supermajority, such as the ones in India and South Africa, does not satisfy
the third feature of the new model. That said, the more weakly a constitution is entrenched against
ordinary majority vote, the closer it is to the new model on a spectrum. By contrast, amendment of a
bill of rights by ordinary majority would satisfy the third feature of the new model.
5
This term has been particularly championed by Janet Hiebert, see supra note 2.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 4
largely indistinguishable from the paradigmatic regime of judicial supremacy of its
neighbor, the U.S.,6
I have not changed my view. I continue to think that the
legislative override power contained in section 33 of the Charter7
is an important and
distinctive structural feature of Canadian constitutionalism despite its infrequent use
and is what clearly places Canada in the new Commonwealth model camp.8
So, on a spectrum in which traditional judicial and legislative supremacy
mark the two poles, the new model can and has taken at least four different forms,
thereby occupying four slightly different intermediate positions. Starting from the
judicial supremacy pole, the first of these is exemplified by the 1982 Canadian
Charter of Rights and Freedoms: (1) a constitutional bill of rights (2) granting the
judiciary power to invalidate conflicting statutes but (3) with a formal legislative
final word in the form of the section 33 override power exercisable by ordinary
majority vote. The second is a statutory or parliamentary bill of rights granting the
judiciary the same power to invalidate conflicting statutes, with a similar legislative
override power. This position is illustrated by the still operative 1960 Canadian Bill
of Rights. The third version is exemplified by the HRA, the Australian Capital
Territory’s Human Rights Act of 2004 (ACT HRA) and the Victorian Charter of
Human Rights and Responsibilities of 2006 (VCHRR): a statutory bill of rights
without the power of judicial invalidation of legislation but instead one new judicial
power to declare statutes incompatible with protected rights that does not affect their
continuing validity, and a second new judicial power (and obligation) to give statutes
a rights-consistent interpretation wherever possible. Both types of judicial decision –
declaratory and interpretive – are subject to the ordinary legal power of the
legislature to have the final word. The fourth variation would be a similar statutory
bill of rights with the second judicial power -- the interpretive one -- but without the
first or declaratory power, as in the original, pre-Moonen, version of the New
Zealand Bill of Rights Act, 1990 (NZBORA).9
A statutory bill of rights alone
without either the interpretive duty or the declaratory power, while certainly a
generally available option, would not be one that satisfies the second requirement of
my definition of the new model and thus, whatever its independent merits, does not
depart from traditional parliamentary sovereignty.
6
Hiebert and Grant Huscroft are among the scholars who have expressed this view.
7
“Parliament or the legislature of a province may expressly declare in an Act of Parliament or of
the legislature, as the case may be that the Act or a provision thereof shall operate notwithstanding a
provision included in section 2 or sections 7 to 15 of this Charter.” Canadian Charter, section 33 (1).
8
In fact, Canada was in this camp even before the 1982 Charter as a result of the 1960 Canadian
Bill of Rights Act, a statutory bill of rights that is still in force. Indeed, as mentioned in the following
paragraph of the text, the CBOR remains as a distinct version of the new model with its (implied)
power of judicial invalidation of statutes.
9
See infra text accompanying notes 63-4.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 5
II. THE BASIC CASE FOR THE NEW MODEL
The core argument for the new Commonwealth model as a whole, it seems to
me, is one of balance. It provides a better balance between two “constitutional
goods”10
or foundational values of modern liberal democratic polities than the
traditional models of constitutionalism by promoting both more equally. These
foundational values are: (1) the recognition and effective protection of certain
fundamental or human rights and (2) a proper distribution of power between courts
and the elected branches of government, including appropriate limits on both. To the
extent that judicial supremacy prioritizes the protection of rights, it does so in a way
that gives too much power to the judiciary at the expense of democratic decisionmaking.
Parliamentary sovereignty prioritizes the democratic decision-making
claims of the political branches at the expense of at least potentially inadequate
protection of rights – in Westminster-based systems typically common law liberties
as supplemented by certain specific statutory safeguards. The new model promises
to recalibrate these two existing, more lopsided options by adequately and effectively
protecting rights through a reallocation11
of power between the judiciary and the
political branches (adding to judicial power if starting from parliamentary
sovereignty and reducing it if starting from judicial supremacy) that denies too much
power to either. As such, it is largely an argument about greater subtlety and balance
in constitutional engineering.
A slightly different way of stating this case is by reference to what has, in
“the latest round of contributions”12
to the debate about judicial review, helpfully
come to be viewed as the two essential issues: (1) are rights better protected with
judicial review; and (2) is judicial review democratically illegitimate?13
In these
terms, the case for the new model is, first, as discussed below, there is a consensus
that rights have been better protected in the various jurisdictions since adopting the
new model and abandoning traditional parliamentary sovereignty, but this same
record also suggests judicial supremacy is not necessary for such protection.
Second, the final legislative word maintains the democratic legitimacy of the lesser,
but still enhanced, judicial powers of the new model. Moreover, one of the best and
most sophisticated recent defenses of the democratic legitimacy of judicial review –
that judicial review’s point of requiring reasonable justification for acts by public
10
I borrow this term from ALAN BRUDNER, CONSTITUTIONAL GOODS (2007).
11
A “reallocation” does not necessarily mean a “transfer” of power from one institution to the
other. Thus, in being given the two new powers of declaring an incompatibility and interpreting
statutes in a rights-consistent way wherever possible, UK courts are not exercising powers previously
held by Parliament. See AILEEN KAVANAGH, CONSTITUTIONAL REVIEW UNDER THE UK HUMAN
RIGHTS ACT 277-278 (2009).
12
Mattias Kumm, Institutionalising Socratic Contestation: The Rationalist Human Rights
Paradigm, Legitimate Authority and the Point of Judicial Review, 1 E.J.L.S. (2009).
13
Among those contributing to this recent crystallization of the issues are RICHARD BELLAMY,
POLITICAL CONSTITUTIONALISM: A REPUBLICAN DEFENSE OF THE CONSTITUTIONALITY OF
DEMOCRACY (2007); Kumm, id.; Wojciech Sadurski, Judicial Review and the Protection of
Constitutional Rights, 22 O.J.L.S. 275 (2002); Jeremy Waldron, The Core of the Case Against Judicial
Review, 115 YALE L.J. 1348 (2006).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 6
authorities is a pre-condition for the legitimacy of law14
– can, I think, be adapted as
a defense of the new model by asking why this “Socratic dialogue” must end with
courts having the final word, rather than permitting the possibility of further
deliberation by a now alerted citizenry and political opposition.15
Indeed, there are
suggestions that the new Commonwealth model may be acceptable to both those
who defend and challenge judicial review. For the former sometimes state that their
defense is neutral as to the form judicial review should take (strong or weak)16
and
the latter that their challenge is only to judicial supremacy or strong-form review.17
Although this article is not primarily about the theoretical justifications for
the new model – which, at least in the relevant countries, are by now fairly wellknown
-- but an assessment of the normatively relevant parts of its practice, let me
put just a little flesh on the bare bones of the previous two paragraphs. My original
article largely focused on the potential benefits of the new model as an alternative to
judicial supremacy that I argued had been adopted (as part of “the American model
of constitutionalism”) and adapted in much of the world after 1945 and again after
1989, rather than as an alternative to the other traditional pole of legislative
supremacy. The main reason for this primary focus was that the pre-enactment
debates, such as they were, in Canada, New Zealand and the United Kingdom
revolved around the choice between a fully constitutionalized bill of rights and some
alternative legalized form. In both Canada and New Zealand, the final bills of rights
emerged after rejection of initial proposals for full constitutionalization along the
lines of the American model and were seen as watered down versions. In the UK,
the HRA was enacted after decades of debate about the merits and, especially, the
constitutional possibility of an entrenched bill of rights, although the specific and
novel legal form of the HRA came as something of a surprise. To a lesser extent, this
focus on the new model versus judicial supremacy was also due to a sense that, in
Mark Tushnet’s words, the Westminster model of legislative supremacy had been
withdrawn from sale.18
14
See Kumm, id.
15
Indeed, Mattias Kumm explicitly suggests that the new model is one way that “Socratic
contestation” might be institutionalized. See Mattias Kumm, Democracy is not Enough: Rights,
Proportionality and the Point of Judicial Review, in M. KLATT ED., THE LEGAL PHILOSOPHY OF
ROBERT ALEXY (2009) (“What deserves a great deal of thought is how to design the procedures and
institutions that institutionalize Socratic contestation….Should judges just have the power to declare a
law incompatible with human rights, leaving it to the legislature to abolish or maintain the law?
[citing The New Commonwealth Model of Constitutionalism]).”
16
See Kumm id.; Richard Fallon, The Core of the Uneasy Case for Judicial Review, 121 HARV. L.
REV. 1693, 1733-34 (2008) (suggesting that the choice between strong and weak forms of judicial
review is a “design question” within the overall institution that he is defending, and that the
desirability of one or the other may depend contextually on the “pathological proclivities” of
particular societies).
17
See Waldron, supra note 13, at 1354 (“The most important difference [among systems of
judicial review] is between what I shall call strong judicial review and weak judicial review. My
target is strong judicial review.”).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 7
The argument for the new model as against judicial supremacy is essentially
that, without a significant sacrifice in terms of rights protection, it produces a better,
more democratically-defensible balance of power between courts and legislatures
with respect to rights. Judicial supremacy, with its associated tendency towards
exclusivity and monologue in rights reasoning, is especially problematic in the
inevitable real world context of reasonable disagreement about the meaning, scope,
application, and permissible limits on the relatively abstract text of a bill of rights
among judges, between courts and legislatures, and among citizens. Moreover, the
second-stage of modern rights adjudication, the stage involving means-end analysis
to determine whether limits on rights are justified, is largely devoid of specifically
legal analysis. The new model permits the rights contained in a legalized charter to
be protected in a less court-centric way that provides a greater role in rights
deliberation for both the political branches and the citizenry. In so doing, it may also
address and help to resolve the well-known problems of (1) the over-legalization or
judicialization of principled public discourse19
and (2) legislative and popular
debilitation that has long been identified as a major institutional cost of
constitutionalization.20
Where legislatures never have final responsibility for rights,
and even more where (as often happens) courts do not take legislative consideration
seriously in their own deliberations, there is an understandable tendency to leave
matters of constitutionality to the judiciary and spend their time on things they do
decide. By giving legislatures the legal power of the final word, the new model
promised to create incentives for such debate. More broadly, the new model
radically and directly dissolves the countermajoritarian difficulty.
Interestingly, the current debate in Australia about whether to enact a national
human rights act is the first to focus squarely and explicitly on the choice between
the status quo of parliamentary sovereignty and some version of the new
Commonwealth model. Accordingly, in this context, the case for the new model as
an alternative to traditional legislative supremacy must be made.
I think this case is that: (1) basic or fundamental or human rights are an
important and valuable way of thinking about collective life in the first place; (2)
they both can and should be more effectively recognized and protected than under
the pre-existing (Westminster-based) system of residual common law liberties
supplemented by certain specific statutory rights within an overall structure of full
parliamentary sovereignty; (3) this more effective protection requires a concise,
legalized statement of affirmative rights;21
and (4) courts have an important, indeed
18
Mark Tushnet, New Forms of Judicial Review and the Persistence of Rights- and DemocracyBased
Worries, 38 WAKE FOR. L. REV. 813, 814 (2003).
19
See, e.g., MARY ANN GLENDON, RIGHTS TALK (199 ); Jeremy Waldron, The Core of the Case
Against Judicial Review, 115 YALE L.J. 1346 (2006); ALEC STONE SWEET, GOVERNING WITH JUDGES:
CONSTITUTIONAL POLITICS IN EUROPE (2000).
20
For an early identification, see JAMES BRADLEY THAYER, JOHN MARSHALL (1901).
21
Here I am contrasting the traditional residual common law conception of rights as whatever the
law does not prohibit with an affirmative conception of rights as presumptive claims against the state.
In particular, I am not referring to the standard distinction between positive and negative
constitutional rights.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 8
essential, role to play in this more effective protection; not necessarily an exclusive
or final one but rather a role that exists alongside, and is ultimately legally
subordinate to, those of the legislative branch of government. The need for greater
protection of rights was certainly felt in the UK prior to the HRA, based on both the
external fact of the UK’s poor record at the European Court of Human Rights
(ECtHR) and the domestic fact of the undermining of civil liberties during the
Thatcher era.22
Part of the reason that rights might be more effectively protected under a
charter is that the legalized dimension of rights discourse that inevitably comes with
a bill of rights (whether statutory or constitutional) is a valuable way of rendering
rights and their limits more concrete and specific, of mooring potentially abstract or
hypothetical issues in reality. Moreover, however vague and indeterminate a bill of
rights may be -- which is part of the reason for not granting judicial supremacy -- as
a form of legalization it is generally far less vague and indeterminate than the
primarily common law liberties of the Westminster-based parliamentary sovereignty
model.
Some form of judicial enforcement power is important to help deal with the
standard concern of certain rights-relevant pathologies or “blind spots” that
legislatures and executives may be subject to, particularly rights claims of various
electoral minorities, legislative inertia based on tradition and convention or resulting
from the blocking power of parties and interest groups, and government hyperbole or
ideology.23
In addition, if (for the reasons already given) granting the judiciary the
legal final word is problematic, this does not necessarily apply to all judicial input
into rights discourse. To the contrary, such input can be helpful as the legally
penultimate word in both informing/spurring rights review by the political branches
both before and after judicial consideration and raising the political costs of
legislative disagreement by alerting the citizenry. In these ways, a judicial role
promises to enhance the quality of legislative rights debate compared to the status
quo.24
Finally, one general advantage of the new model over both of the traditional
ones is that it has greater potential to actively involve all three branches of
government in rights review and creates a broader rights consciousness among the
citizenry: what Grant Huscroft has helpfully referred to as a “ground up” rights
culture rather than the typically “top-down” one of judicial supremacy.25
So in
Canada, New Zealand, the UK and Australia, the new model has not only created
new powers of judicial involvement in rights enforcement but mandated new forms
22
The classic work here is KEITH EWING AND CONOR GEARTY, FREEDOM UNDER THATCHER
(1990).
23
See, e.g., ALEXANDER BICKEL, THE LEAST DANGEROUS BRANCH (1962); Rosalind Dixon, The
Supreme Court of Canada, Charter Dialogue, and Deference, 47 OSGOODE HALL L. J. 235 (2009);
Kumm, supra note 15; Michael J. Perry, Protecting Human Rights in a Democracy: What Role for
Courts?, 38 WAKE FOR. L. REV. 635 (2003).
24
See Gardbaum, supra note 1, at 744-48; Perry, id.
25
Grant Huscroft, Constitutionalism from the Top Down, 45 OSGOODE HALL L.J. 91 (2007).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 9
of pre-enactment rights review within the legislative process.26
A second advantage
is that the new model transcends the existing either/or choice between a legal and a
political conception of rights by providing conceptual and normative space for both.
Of course, the new model is not without its own possible costs and
disadvantages compared to parliamentary sovereignty. Like the benefits, these are
basically of two kinds: normative and empirical.27
Very briefly, the normative
concern is that there are basic democratic objections to any degree of judicial power
over legislation beyond traditional presumptions and techniques of statutory
interpretation. The empirical/practical are (1) that there is no need to protect rights
more or in a different way, the status quo is fine, (2) that rights are not in fact better
protected under the new model and/or (3) that the new model is unstable -- the
theoretical balance of the intermediate position does not exist or cannot be
maintained in practice but inevitably leads to the excesses of judicial power and
rights talk commonly associated with strong systems such as the US. As we will see,
these latter two claims but particularly the final one, have been commonly made
criticisms of how the new model has worked in practice.
III. THE CRITERIA OF SUCCESS (OR FAILURE)
In order to be in a position to evaluate the new model in practice, it is first
necessary to clarify what the relevant criteria of success are. These, I think, follow
fairly straightforwardly from the basic case for the new model. First, is there
effective, or at least adequate, protection of the rights recognized in the new charters
or bill of rights? Second, has the appropriate reallocation of power between courts
and the political branches with respect to rights – including limits on both -- taken
place not only de jure but also de facto so that they are now in practice in better
balance? This second criterion raises related questions about both the stability (is it
reverting towards one or other of the two poles?28
) and the practical distinctiveness
(versus the legal or formal) of the new model. Where the intended switch is from
parliamentary sovereignty as traditionally conceived – the precise question here is
whether (1) the right amount of power, (2) too little, or (3) too much has been given
to (or taken by) the courts. Too little power would suggest the continuing dominance
26
Indeed, Janet Hiebert has argued that such political rights review (i.e., parliamentary
consideration and interpretation of rights before rather than after judicial review) is what “most
distinguishes rights practices in Canada and the UK from the US.” Janet Hiebert, New Constitutional
Ideas: Can New Parliamentary Models Resist Judicial Dominance When Interpreting Rights?, 82
TEX. L.REV. 1963 (2004).
27
These potential costs have recently been well-canvassed in James Allan, The Victorian Charter
of Human Rights and Responsibilities: Exegesis and Criticism, 30 MELBOURNE UNIV. L. REV. 880
(2006).
28
See Mark Tushnet, Weak-Form Judicial Review: Its Implications for Legislatures, 23 SUP. CT.
L. REV., 213, 224 (2004), and also supra note 15, for general or theoretical reasons to fear such
instability in either direction. As we will see, this claim of instability has been the basis for many
critiques of the actual practices in each country.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 10
of traditional parliamentary sovereignty; too much would suggest a change to
something not far removed in practice from judicial supremacy.
These criteria are admittedly rather abstract. To make them more concrete, it
might be helpful to outline the ideal working of the new model – in the sense of what
we would actually want to see in practice -- in light of the basic normative argument
for it.
The deliberate institutional engineering of the new model suggests that there
are three stages at which the rights protected under the relevant charter come into
play. The first stage is pre-enactment political rights review. I fully concur with
Janet Hiebert that such legally mandated review by the government is an important,
attractive and distinctive part of the new model and its effective protection of rights,
although I don’t believe that this is what “most distinguishes” the new model from
judicial supremacy and U.S.-style constitutionalism.29
Rights review and
deliberation by all three branches of government (including, importantly, legislative
oversight of the executive) and not just by the courts is critical to the success of the
model and to ensuring rights consciousness from the “ground up.” In the case of
legislation, this requires political rights review before enactment and judicial
involvement, as well as after. Moreover, this stage should not only be about when
the political branches engage in review but what that review consists in. In other
words, executive and legislative review should not be exclusively legal in nature,
focused either on reasonable interpretive pluralism within the law or predicting what
the courts will ultimately do, but should bring a broader, freer perspective of
principle to the issue than is typical of judicial reasoning.
The second stage of the new model is legal rights review by the courts under
their respective new powers. Here, two factors should distinguish, at least in part,
judicial review under the new model from that under judicial supremacy. First,
because of the intended joint, or three-branch, responsibility for rights that is one of
the potential advantages of the new model, courts should take seriously the political
rights review at stage one and not ignore or treat contemptuously as a usurpation of
judicial supremacy the rights deliberations of the political branches, as sometimes
occurs in US-style systems.30
This judicial posture also properly acknowledges the
reality and inevitability of reasonable disagreement about both rights and their limits
in the bill of rights that is one of the major reasons for the new model (again, at least
as compared with judicial supremacy) in the first place. Obviously, to the extent that
the political rights review at the first stage approaches the ideal version suggested
above, the basis for this judicial posture towards it is further strengthened. Second,
and notwithstanding the first, given the legal final word that is bestowed on the
political branches, the political community as a whole and its elected representatives
will benefit most from the best judgment of the judiciary on the merits; that is, an
independent consideration that takes into account but is not foreclosed by the views
29
See Hiebert, supra note 19, at 1985.
30
For an example of such alleged usurpation, see City of Boerne v. Flores, 521 U.S. 526 (1997).
This critique of judicial supremacy, among others, is made in Jeremy Waldron, Some Models of
Dialogue Between Judges and Legislators, 23 SUP. CT. L. REV., 7, 39-46 (2004).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 11
of the political branches expressed at stage one. In this way again, the nature and
function of “penultimate judicial review” within the new model differentiate it from
judicial supremacy31
– in which the judiciary has full practical responsibility for the
final decision and its consequences.
The third stage is the possible exercise of the legislative final word. Here the
ideal of the new model, I suggest, is principled and focused legislative debate
following a judicial finding of invalidity (Canada32
) or incompatibility (elsewhere)
on whether and how to amend/repeal the statute at issue, in which the legislature -- in
a similar but inverse way from the second stage -- takes seriously the judicial view
but does not automatically defer to it. A similar process is to be encouraged after
exercise of the interpretive duty where the legislature contemplates overruling the
court’s interpretation of a statute as mistaken. Where this occurs and the result is
within the bounds of reasonable disagreement with the courts, legislative exercise of
its legal power should generally be considered politically legitimate. I think the
process here is the most important thing and not the outcome – so that principled and
serious legislative consideration resulting in decisions to comply with the courts
manifest what the new model seeks to achieve as much as decisions not to comply,
as long as the latter is generally taken to be a realistic political possibility. In other
words, compliance per se is not a problem although a “culture of compliance” is.33
Now that we have some concrete sense of what we are looking for, we can
turn to the main task of this article and apply these criteria to the experience of the
countries at issue, each of which has made the move from traditional parliamentary
sovereignty.
IV. ASSESSING THE CANADIAN, NEW ZEALAND AND UK EXPERIENCES
Although a good deal has been written in the last few years on the workings
of the new model, most assessments have been country-specific rather than systemic
or comparative.34
Moreover, some of this commentary gives the impression that its
authors are locked into initial positions of support or opposition to change – to the
Charter, the NZBORA or the HRA – and have a tendency to assess subsequent
experience through this lens.
With these two qualifications in mind, I think it is fair to say that overall,
academic judgments have not been particularly favorable. Although far from
uniform, perhaps the predominant line of criticism is that, whatever the advantages
in theory, in practice the new model as a whole – especially in Canada and the UK --
has not proven itself to be as different from US-style judicial review as was claimed
31
See Perry, supra note 16.
32
Under Section 33, Canadian legislatures may also override rights preemptively; i.e., immunize
them against future judicial review.
33
Hiebert uses this term to describe what she views as the subservience of Canadian and UK
legislatures to the courts. See Hiebert, supra note 2, at 28.
34
Hiebert and Tushnet are among the exceptions.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 12
or hoped but has rather evolved into something barely distinguishable from it.35
In
presenting my own comparative evaluation, I consider each country in turn before
attempting to make and justify general observations and assessments.
A. Canada
Under the Charter, Canada has a constitutionalized and entrenched bill of
rights with full judicial review powers. What formally and most obviously
distinguishes the Canadian system from the standard model of judicial supremacy is
the rejection of its final prong: that judicial decisions enforcing the supreme law bill
of rights cannot be overruled by ordinary legislative majority – but only by the
supermajority requirement of the formal amendment process. This rejection, of
course, takes the form of the “notwithstanding” clause of section 33, permitting
federal and provincial legislatures to declare by ordinary majority vote for a
renewable period of five years that a statute “shall operate notwithstanding a [rights]
provision included in section 2 or sections 7 to 15 of this Charter.”36
At least among academics, the overall debate on the success of the Charter in
practice within Canada has been dominated by (in my terms) the balance of power
issue – has too much power, too little, or the right amount been granted to the courts
relative to legislatures. It is not seriously questioned that in general terms rights are
now better protected than before 1982 although, of course, there is the expected
disagreement about both the existence and scope of certain particular rights and how
adequately others are protected.
The context in which the balance of power debate has taken place is the
uncontested fact that section 33 has largely fallen into non-use, the provincial
legislatures and especially the federal parliament have been mostly unwilling to
challenge judicial decisions on rights. How much this is due to (1) the particular
history of perceived initial illegitimate use of section 33 by Quebec, (2) to the
relatively non-controversial and restrained nature of the Supreme Court of Canada’s
(SCC) exercise of its judicial review powers, (3) to the particular wording of section
33 as seemingly involving “rights-misgivings” rather than “rights disagreements”37
on the part of legislatures, or (4) the inherent difficulty of (at least, reactive, if not
preemptive) use where a legislature is affirmatively required to overrule a court
decision, is a matter of disagreement. In any event, the virtual dormancy of section
33 has led some commentators to claim that Canada currently has a system of
judicial supremacy that in practice is as strong as and essentially indistinguishable
from the paradigmatic example of the United States.38
In this way, it is argued, the
35
See, e.g., Hiebert, supra note 19; James Allan, You Read Words In, You Read Words Out, You
Take Parliament’s Clear Intention and You Shake It All About, ssrn abstract number 1443626.
36
Although it is sometimes suggested that legislatures do not have the legal power of the final
word under section 33 because when an override expires at the end of five years the original court
decision is reinstated at that point, in my view the fact that the override is renewable (every five years)
means that the legal power of the final word remains with the legislature.
37
See Waldron, supra note 19.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 13
hoped-for differences and benefits of a distinctive Canadian approach to
constitutionalism have not materialized: a cautionary tale about enhancing judicial
power generally and granting power to invalidate legislation in particular.
By contrast, other Canadian commentators – including Peter Hogg and his
co-authors, and Kent Roach -- have rejected this interpretation of Charter experience
and argued that, even absent use of section 33, Canadian judicial review is different
from and weaker than that of the US. This is because, due to certain other structural
features of the Charter (primarily section 1, the general limitations clause),
legislatures in practice usually have the final word in that they can and do re-enact
laws invalidated by the courts, so permitting them to achieve their purposes by other
means.39
In this way, the claim is that there is de facto legislative, not judicial,
supremacy. Hogg et al first referred to this phenomenon of successful legislative
sequels as “dialogue” between courts and legislatures – and their initial 1997 article
spawned a new cottage industry in Canada under the label “dialogue theory,” which
has taken on a life of its own and become the predominant theoretical approach to
the Charter. It has since spread to the UK and Australia where the new model as a
whole is often referred to as “the dialogue model”40
and/or justified on the basis that
it promotes “democratic dialogue.”41
My own assessment is slightly different than both. As I’ve already said,
contra the “as strong as the US thesis,” I think the existence of section 33 is the
distinctive feature of the Charter by comparison with other systems with
constitutional bills of rights and the judicial power to invalidate statutes, and is what
makes it an instance of the new model. Apart from a limited borrowing by Israel
under some but not all of its constitution-by-installments, the Knesset-enacted Basic
Laws, no other country has an equivalent constitutional provision to section 33.42
Even now, it makes a practical difference because were the SCC to decide a case that
triggered the degree of controversy marked by certain judicial decisions in other
countries with strong-form judicial review, there is a decent chance that the relevant
legislature would seriously consider and perhaps use the override. Indeed, Alberta
38
These include Janet Hiebert, supra note 2, and Grant Huscroft, supra note 18. Of course, there
are other Canadian commentators, those who oppose section 33 and support judicial supremacy, who
applaud this development. Cites.
39
See Peter W. Hogg and Alison A. Bushell, The Charter Dialogue Between Courts and
Legislatures (Or Perhaps The Charter of Rights Isn’t Such A Bad Thing After All), 35 OSGOODE HALL
L.J. (1997); Peter W. Hogg, Alison A. Bushell Thornton & Wade K. Wright, Charter Dialogue
Revisited – Or “Much Ado About Metaphors,” 45 OSGOODE HALL L.J. 1 (2007); KENT ROACH, THE
SUPREME COURT ON TRIAL: JUDICIAL ACTIVISM OR DEMOCRATIC DIALOGUE (2001); Kent Roach,
Dialogic Judicial Review and its Critics, 23 SUP. CT. L. REV. 49 (2004); Kent Roach, Sharpening the
Dialogue Debate: The Next Decade of Scholarship, 45 OSGOODE HALL L.J. 169 (2007).
40
See the Australian debate; for example, the National Consultation Committee’s Report
recommending adoption of “the dialogue model” by enacting a national human rights act.
41
See, e.g., ALISON YOUNG, PARLIAMENTARY SOVEREIGNTY AND THE HUMAN RIGHTS ACT 115-
143 (2009); Tom Hickman, Constitutional Dialogue, Constitutional Theories and the Human Rights
Act 1998, 2005 PUB. L. 306.
42
A version of section 33, but importantly with a two-thirds majority requirement, was adopted in
Poland between 1986 and 1997 and Romania until 2003, and the VCHRR contains a statutory version
in section 31.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 14
came very close to using the override in response to the SCC’s decision in Vriend43
--
reading into its provincial human rights statute sexual orientation as a prohibited
basis of discrimination when the legislature had expressly decided to omit it – but
ultimately did not; although Alberta did subsequently use the override preemptively
in enacting its 2000 statute banning same-sex marriage: the Marriage Amendment
Act. In fact, according to a well-regarded article, section 33 has been used more
often than generally appreciated, 16 times from 1982 to 2001.44
The separate point
that the conduct of Canadian courts is impacted by a self-perception as exercising
their powers in the shadow of section 33 in a way that constitutional courts in
countries with de jure judicial supremacy do not, even if that shadow is smaller than
originally expected, may also be true but is difficult to establish counterfactually.45
For these reasons, I believe that its importance as a legally authorized outlet for
popular or legislative disagreement with the court is not “purely” formal, is not
simply and totally negated by the current general norm of non-use. Section 33 is not
(yet) the equivalent of the Royal Assent.
So I think the marginalizing of section 33 as a distinctive structural feature of
the Charter by this first group of commentators is slightly overdone. This is certainly
not to deny either that the general norm of non-use seriously undermines any
potential claim that the Charter version of the new model – placing the burden of
inertia on the legislature in exercising the power of the final word – has been the
most successful or, as Hiebert among others has argued, that steps might usefully be
taken to try to ease its use.46
If not already too late, these could include attempting to
reframe popular conceptions of section 33 from “overriding” or disregarding of
rights to permitting reasonable interpretive disagreement about what rights there are
– perhaps by amending the wording of section 33 to reflect this, along the lines of the
VCHRR version.47
On dialogue theory generally, I cannot help but feel that the initial very
specific but metaphorical use of the term “dialogue” by Hogg et al (to refer to
successful legislative sequels) has unwittingly led much of the academic debate –
43
Vriend v. Alberta, [1998] 1 S.C.R. 493.
44
Tsvi Kahana, The Notwithstanding Mechanism and Public Discussion: Lessons from the
Ignored Practice of Section 33 of the Charter, 44 CAN. PUBLIC ADMIN., 255 (2001).
45
Perhaps R. v. Sharpe, [2001] 1 S.C.R. 45, a case in which, rather than striking down the statute,
the SCC read down the scope of the offence of possession of child pornography even when it was
clear that Parliament had intended the offence to cover such matters, and found the reduced scope of
the statute a permissible limitation on free expression under section 1 – and the somewhat similar case
of R. v. Butler, [1992] 1 S.C.R. 452 – provide the best examples of the SCC operating under the
“shadow” of section 33. But cf. Sujit Choudhry and Claire Hunter, Measuring Judicial Activism on
the Supreme Court of Canada, 48 MCGILL L. REV. 3 (2003). On the other hand, it has also been
suggested, by me for one (see infra TAN 156), that courts exercising the power of penultimate, rather
than ultimate, judicial review might be expected to be bolder in their interpretation and application of
rights, freed as they are from the full practical responsibility for their rights decisions.
46
See Janet Hiebert, Is it Too Late to Rehabilitate Canada’s Notwithstanding Clause, 23 SUP. CT.
L. REV. 169 (2004); Waldron, supra note 19; Jeff Goldsworthy, Judicial Review, Legislative Override,
and Democracy, 38 WAKE FOR. L. REV. 451 (2003).
47
Section 36, VCHRR.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 15
both inside and outside Canada -- up something of a blind alley. Dialogue theory has
come to dominate Canadian discussion of the Charter,48
with dialogue often viewed
(1) more literally than Hogg intended, (2) as an end in itself, (3) as the exclusive or
predominant criterion for characterizing and evaluating the Charter and, increasingly,
the new model as a whole, and/or (4) as constituting a distinct form of judicial
review (“dialogic”)―rather than one convenient way among others of referring to or
specifying the new model's promise of greater institutional balance of power. I think
the notion that what distinguishes Canada, and perhaps the new Commonwealth
model as a whole, is dialogic judicial review is problematic because on Kent Roach’s
definition -- any system in which rights in a bill of rights can be limited or
overridden by ordinary legislation49
-- there really is no “non-dialogic review”
anywhere, given the very many other countries with systems of judicial supremacy
and express limits on rights applied via the proportionality test, but not even in the
US. In other words, in the sense used by Hogg and Roach in the Canadian context,
dialogue is an overinclusive term and insufficiently distinctive of the new model as a
whole.
Moreover, in a broader, less technical sense of the term, in various forms and
to varying degrees, dialogue between various actors and institutions occurs in all
types of legal and political systems, including abstract review by constitutional
courts and US-style judicial review,50
so it is neither equivalent nor exclusive to the
new Commonwealth model. Overall I fear that, while general use of the term was
originally one among several helpful ways of characterizing and justifying the new
model (I employed it this way myself), dialogue theory as it has developed and
mushroomed threatens to blur the distinctions between the new Commonwealth
model and judicial supremacy. Unless and until there is far greater specificity and
precision, in the form of a customized normative theory of when courts and,
especially, legislatures should exercise their respective powers under the new
model,51
“dialogue” will remain too generic and amorphous a concept to serve as
either (1) the ultimate criterion of institutional balance or (2) the major benefit or
justification of the new model, as compared to a claim such as providing the least
restriction on democratic-decision-making consistent with an adequate protection of
basic rights.
Turning specifically to the Hogg/Roach thesis about de facto legislative
supremacy, I generally agree with it that judicial review is weaker in Canada than in
the US. I am less certain, however, that this is due primarily to the existence of
48
Look in the index of any book on the Canadian Charter and you will find long and multiple
entries under “dialogue.”
49
Kent Roach, Dialogic Judicial Review and its Critics, 23 SUP. CT. L. REV. 49, 55 (2004).
50
As Barry Friedman has exhaustively charted between the judiciary and public opinion. See
BARRY FRIEDMAN, THE WILL OF THE PEOPLE (2009).
51
Recent promising attempts in this direction include ALISON YOUNG, supra note 40 (proposing
in the UK context a democratic dialogue model of reasonable interpretive disagreement as the
normative criterion for use of the declaration of incompatibility) and Rosalind Dixon, see supra note
23 (proposing a new theory of dialogue in the Canadian context in which the SCC should defer to
reasonable legislative sequels in “second look” cases).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 16
section 1, the general limitations clause, as a structural feature of Canadian
constitutionalism. As already discussed, I think section 33 is the critical and
distinctive structural feature of the Charter from a general constitutional and
democratic theory perspective and not section 1, which (a) has express equivalents in
many modern systems of judicial supremacy around the world,52
and (b) even in the
US, courts have long implied limits into almost all seemingly absolute textual and
other rights.53
Apart from section 33, Canadian judicial review is probably weaker
than that of the US in practice but arguably less because of the existence of section 1
or limits on rights per se than because of the proportionality test that the SCC
subsequently implied into it and, in particular, the relatively deferential way the SCC
applies that test54
– at least as compared with (a) the strict and intermediate scrutiny
standards of review for limits on rights in the US and (b) how the proportionality test
is applied by most other constitutional courts.55
The SCC’s suspended declaration of
invalidity as a remedy creating strong incentives for a legislative response is, as Kent
Roach argues, another concrete dialogue-inducing feature of the Charter.56
However,
it is neither one without equivalents in systems of judicial supremacy57
nor one that
permits the legislature an independent choice of final word.
To the extent it has a comparative dimension, Canadian discussion of the
Charter tends to be a little fixated on the US, as has perhaps been true of Canadian
constitutionalism throughout its history.58
But from a broader comparative
constitutional context whether, using Hogg’s criterion of successful legislative
sequels (and thus putting section 33 aside), Canada has a weaker system of judicial
review not just than the US but than other countries with a judicial invalidation
power and the proportionality test for limits is less clear. If not, this is unlikely to
rebut the critics’ claim that Canada has devolved into de facto judicial supremacy
and strong review. If so, it is likely the result not of any structural or formal feature
of the Charter – and not any new model feature59
-- but rather the SCC’s current and
52
General or special limitations clauses are part of what Lorraine Weinrib has characterized as the
“postwar paradigm.” Lorraine Weinrib, The Postwar Paradigm and American Exceptionalism, in
SUJIT CHOUDHRY ED., THE MIGRATION OF CONSTITUTIONAL IDEAS (2006).
53
See Stephen Gardbaum, Limiting Constitutional Rights, UCLAL. REV. (2007)
54
See Sujit Choudhry, So What is the Real Legacy of Oakes? Two Decades of Proportionality
Analysis under the Canadian Charter’s Section 1, (2006), 34 S.C.L.R. (2d).
55
See Gardbaum, supra note 45.
56
See Roach, supra note 42.
57
Abstract judicial review in centralized systems has a generally similar, constrained dialogueinducing
feature of setting constitutional parameters within which legislatures can make their choices.
The South African constitution expressly grants the power of suspended invalidity to the courts and
the German Constitutional Court has developed a range of equivalent techniques, including finding
statutes unconstitutional but not void (“unvereinbar” versus “nichtig”) -- meaning they are in force
during a transition period pending correction by the legislature -- and upholding statutes but warning
they will be invalidated in future if not amended or repealed.
58
See Jamie Cameron, The Charter’s Legislative Override: Feat or Figment of the Constitutional
Imagination?, 23 SUP. CT. L. REV. 135 (2004).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 17
controversial practice, in exercising its judicial review power, of relative deference to
the government in applying the three prongs of the proportionality test – a judicial
practice that (however much one might independently support it on democratic
grounds60
) a future SCC majority might easily change.
In sum, I view Canada as part of the new model because of section 33, rather
than section 1, but as suffering from a serious practical problem due to the apparent
near-convention against its use. The problem is less that the override power is rarely
exercised per se than that this near convention seems largely to exclude the sort of
political discussion about rights that the ideal working of the new model calls for.
B. New Zealand
Following rejection of the White Paper proposing a fully constitutionalized
bill of rights, New Zealand settled in 1990 for a statutory one, although it is different
in one crucial respect from what Paul Rishworth has described as the “antiprecedent”
of the 1960 Canadian Bill of Rights.61
Briefly, the New Zealand Bill of
Rights Act (NZBORA) contains a relatively narrow set of civil and political rights by
comparative standards derived (selectively) from and explicitly stated to affirm New
Zealand’s commitment to the ICCPR,62
together with a general limitations clause
borrowed from section 1 of the Canadian Charter.
These statutory rights are protected in two ways. First, executive and
legislative pre-enactment rights review is fostered by requiring the Attorney-General
to scrutinize all bills introduced into the House of Representatives and to bring to its
attention any that appear incompatible with the rights.63
Second, judicial protection
is afforded by requiring courts to interpret statutes so that, according to section 6,
“wherever an enactment can be given a meaning that is consistent with the rights and
freedoms contained in this Bill of Rights, that meaning shall be preferred to any
other meaning.” Section 4, however, expressly denies the courts power to invalidate
or deny operative effect to any statute – previous or subsequent to NZBORA -- that
is inconsistent with any protected right.64
This is the one crucial difference from the
CBOR, which the SCC had held to impliedly authorize it to invalidate irredeemably
59
Although I suppose it is conceivable that this practice itself owes something to the shadow of
section 33.
60
See, e.g., Gardbaum, supra note 45, at 830-852.
61
Paul Rishworth, The Inevitability of Judicial Review under “Interpretive” Bills of Rights:
Canada’s Legacy to New Zealand and Commonwealth Constitutionalism?, in id.
62
As Andrew and Petra Butler note, there is a lot on criminal procedure and, for example, nothing
on education, privacy, or property rights. See ANDREW S. BUTLER AND PETRA BUTLER, THE NEW
ZEALAND BILL OF RIGHTS ACT: ACOMMENTARY 4 (2005).
63
NZBORA, section 7.
64
“No court shall, in relation to any enactment (whether passed or made before or after the
commencement of this Bill of Rights), -- (a) Hold any provision of the enactment to be impliedly
repealed or revoked, or to be in nay way invalid or ineffective; or (b) Decline to apply any provision
of the enactment – by reason only that the provision is inconsistent with any provision of this Bill of
Rights.” NZBORA, section 4.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 18
inconsistent statutes, although only once did it actually use that power.65
In this
sense, NZBORA creates an interpretive bill of rights rather than an overriding one,
and the legislative final word lies in the legal power to enact a statute that cannot be
interpreted consistently with the rights or to amend an existing one – whether already
interpreted by the courts or not -- to this effect. NZBORA does, however, grant the
courts power to invalidate executive action violating the protected rights, unless such
action is mandated by legislation―and this has been by far the most active area of
NZBORA litigation, especially with regard to police action.66
Since 1990, the NZBORA has grown from essentially an afterthought to an
instrument of constitutional stature67
despite its ordinary law status,68
with a
correspondingly growing impact on the workings of all three branches of
government. Early on the NZ Court of Appeal declared that a “generous” and
“purposive” approach to rights interpretation was appropriate under NZBORA,
adopting the language of the SCC under the Charter.69
More concretely, in perhaps
the boldest of all judicial action under NZBORA, it has undergone two outwardly
significant changes both remedial in nature. First, despite silence on the subject in
NZBORA, the Court of Appeal created the public law remedy of damages for
executive violation of a protected right, even in the face of a separate statute
seemingly immunizing the relevant government body from liability.70
Second, in
2000, despite absence of such an express power, the Court of Appeal in dicta
strongly suggested, based on inference from statutory text and structure, that it had
formal authority to “indicate” or “declare” that statutes are inconsistent with
NZBORA, although such statutes remain fully valid and applicable.71
This authority,
however, has not yet clearly been exercised and indeed the courts have “resisted
endorsing it in later cases.”72
Paul Rishworth has argued, in general agreement with
the Court’s rationale, that notwithstanding the express rejection of a judicial power to
invalidate statutes in section 4, the power to indicate inconsistency is inherent or
embedded in the structure of the NZBORA or, indeed, any statutory bill of rights that
requires interpretive solutions where possible but retains the idea that such solutions
may not be possible because of irremediable inconsistency. As he puts it: “The very
65
This was in the same case as the power was implied: R. v. Drybones, [1970] S.C.R. 282. In
most other respects, however, the SCC is generally thought to have underenforced the CBOR
66
Although not stated in so many words, this power has been taken to follow from the facts that
(a) NZBORA applies to the legislative, executive, and judicial branches of government but (b) only
legislation inconsistent with the bill of rights is stated still to be valid and effective.
67
See Andrew & Petra Butler, supra note 54, at 9.
68
Indeed, because of the express ouster of the normal doctrine of the implied repeal of an
inconsistent earlier statute by NZBORA contained in section 4, NZBORA in this sense has less-than
ordinary statute status.
69
Ministry of Transport v. Noort [1992] 3 NZLR 260 (CA).
70
Simpson v. Attorney-General [Baigent’s Case] [1994] 3 NZLR 667 (CA).
71
Moonen v. Board of Film and Literature Board [2000] 2 NZLR 9 (CA).
72
Andrew & Petra Butler, supra note 54, at 1026.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 19
process of determining that one is bound, by section 4, to apply an inconsistent
enactment is itself a declaration of that enactment’s inconsistency.”73
As we will see,
in both of these ways, the Court of Appeal brought NZBORA into line with the
express provisions of the UK’s Human Rights Act, enacted in 1998, thereby ending
the most obvious structural differences between the two.74
In terms of the Attorney-General’s duty to report, there have been
approximately 40 instances of reports of apparent inconsistencies,75
which is very
different from the situation in both Canada and the UK where there have been zero
and one respectively under equivalent duties.76
Like Canada, but unlike the UK,
there is no specialized parliamentary committee that considers these reports and
other evidence of a bill’s compatibility with NZBORA. In terms of the judicial
power, despite the analytical clarity between sections 4 and 6 – where possible,
rights-consistent interpretation of statutes are to be given but where not, courts will
apply the inconsistent provision -- in practice there is inevitably a fine line between
whether a rights-consistent interpretation “can” or cannot be given, depending in part
on what this undefined term is taken to mean, and so an area of subjective judgment
as to which section to employ in a give case. Section 4, of course, expresses the core
value of legislative decision-making and section 6 the value of rights protection.
How have judges drawn this line and, in particular, to what extent have they used the
section 6 mandate aggressively or radically to essentially modify rather than interpret
statutes? Again, as we will see, this issue has been an important aspect of the British
experience under the HRA with a similar power. Indeed, this interplay between the
interpretive duty of courts to give statutes rights-consistent readings whenever
possible on the one hand and the ban on courts from denying inconsistent statutes
operative effect on the other is the most complex and controversial aspect of both
NZBORA and the UK HRA in practice – do aggressive or strained exercises of the
former violate the latter?
The general sense among New Zealand commentators is that this power/duty
has been used overall with reasonable moderation,77
without the types of “judicial
vandalism”78
or “interpretation on steroids”79
that (as we will see) have been alleged
in the UK. The leading example often given of such restraint is the refusal of the
73
Rishworth, supra note 53, at 266.
74
The major remaining difference is the contextual factor that UK courts work within large
shadow of the ECtHR, whereas NZ courts have no such equivalently effective international oversight.
75
Grant Huscroft, Reconciling Duty and Discretion: The Attorney General in the Charter Era,
available at http://ssrn.com/abstract=1326706.
76
The Communications Act, 2003.
77
See PAUL RISHWORTH, GRANT HUSCROFT, SCOTT OPTICIAN, RICHARD MAHONEY, THE NEW
ZEALAND BILL OF RIGHTS 143-147 (2003); Andrew Butler & Petra Butler, supra note 54.
78
R (on the application of Anderson) v. Secretary of State for the Home Department, [2002]
UKHL 46 (“to do so would not be judicial interpretation but judicial vandalism,” Lord Bingham at
para. [30]).
79
See Allan, supra note 24, at 881.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 20
Court of Appeal in Quilter (1998) to interpret the gender-neutral language of the
1955 Marriage Act as embracing same-sex marriage, despite not only section 6 but
also section 19 NZBORA, as amended by the 1993 Human Rights Act to prohibit
discrimination on grounds of sexual orientation.80
The fact that six years after
Quilter, the House of Representatives enacted the Civil Unions Act after serious
parliamentary debate has also been posited as evidence that judicial-legislative
dialogue is working reasonably well.81
Moreover, there is at least one pair of cases
in which NZ and UK courts came to opposite conclusions in interpreting essentially
identical statutory language imposing a reverse onus of proof on criminal drug
defendants, with the UK court giving it the more rights-consistent but arguably
strained meaning and the New Zealand court refusing to do so and relying instead on
section 4.82
And two years ago, in affirming this original 1991 New Zealand
interpretation and rejecting the subsequent British one, the new New Zealand
Supreme Court expressly and self-consciously discussed and rejected the “broader”
use of the equivalent and textually similar power by UK courts in favor of the
narrower requirement of “reasonably possible” meanings (versus what Tipping J.
characterized as “unreasonably possible” ones in the UK83
).
On the other hand, although no New Zealand case has gone quite so far as the
most “adventurous”84
of the UK ones in interpreting statutory provisions in ways that
Parliament seemingly did not intend, some New Zealand judges have espoused more
expansive statements of the judicial power to give merely possible rather than
plausibly intended meanings to statutes, and there are dicta to this effect in majority
opinions in a few cases.85
Thus, in Poumako, although the case was decided on other
grounds, the majority opinion stated: “It is not a matter of what the
legislature…might have intended. The direction [in section 6] is that wherever a
meaning consistent with the Bill of Rights can be given, it is to be preferred.”86
As
one commentator concludes: “There is little consistency of approach or application
[to section 6]…One can find support in the case law for almost any view relevant to
s6….In this respect, [NZ]BORA case law is no different from the case law that has
developed in respect of s3(1) of the HRA (UK).”87
80
Quilter v. Attorney-General [1998] 1 NZLR 523 (CA).
81
Petra & Andrew Butler, 16 Years of the New Zealand Bill of Rights (manuscript on file with
author).
82
R v. Phillips [1991] 3 NZLR 175 (CA), affirmed in R v. Hansen [2007] NZSC 7 (SC); cf R. v.
Lambert [2001] UKHL 37.
83
Hansen, id., at 158.
84
Andrew & Petra Butler, supra note 54, at 345.
85
Flickinger v. Crown Colony of Hong Kong [1991] 1 NZLR 439 (CA); R v. Poumako [2000] 2
NZLR 695 (CA); R v. Pora [2001] 2 NZLR 37 (CA).
86
Poumako, id.
87
Andrew & Petra Butler, supra note 54, at 183.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 21
Moreover, according to Paul Rishworth, only a bare handful of cases have
involved “considered” reliance on section 4 (i.e., applying the statute after it has first
been found to be an unreasonable limit on, and so inconsistent with, the bill of
rights),88
although there are several more cases of “unconsidered” or blanket reliance
on section 489
-- and there has been no clear use of the implied power to formally
declare inconsistency.90
This compares with the 15 surviving uses of section 4 and
the declaratory power in the UK during the same time period of 2000-2009. Overall,
there is perhaps only a small difference in general approach and application of the
interpretive power between the two countries – a slightly weaker presumption in
New Zealand than in the UK that Parliament intends to act consistently with rights.
In sum, there appears to be broad consensus among leading New Zealand
commentators that NZBORA is working reasonably well and as anticipated, that
criminal procedure and freedom of expression rights have been adequately if not
boldly protected while protection of other rights has been more minimal, that New
Zealand is better of with NZBORA than without but that there is no need or
justification for moving to a fully constitutionalized bill of rights.91
This consensus
includes the perception that the legislature still decides basic issues rather than the
courts, and perhaps also that – due to personnel changes -- the courts have recently
become more conservative and less “generous” in the application of NZBORA.
There is some disagreement, however, about the exact balance between judicial and
legislative power, and how it might be improved. Thus, one “section 4 proponent”
has expressed concern that the implication of the declaratory power heralds the
further erosion of constraints on judicial power that have marked the new model
elsewhere.92
By contrast, two other prominent commentators have argued that the
balance needs to be recalibrated in favor of rights protection by repealing section 4
and replacing it with a section 33-type override mechanism, thereby ensuring that
legislative inertia works for rather than against rights claimants who otherwise have
too little incentive to bring proceedings.93
This broad academic consensus is a little
startling in comparison with its visible absence in Canada and the UK, but I have no
good reason to think it does not bear some relationship to the reality.
88
Paul Rishworth, Interpreting Enactments: Sections 4, 5, and 6, in RISHWORTH ET AL., THE NEW
ZEALAND BILL OF RIGHTS, supra note 51, at 156-7.
89
The Butlers record 18 cases in total as of December 2005, supra note 58, at 191. Indeed, contra
Rishworth, they claim that “reliance on Section 4 has been substantial.”
90
Rishworth, supra note 51, at 156-7.
91
In addition to the sources cited in supra note 51, see also Michael Taggart, Tugging on
Superman’s Cape, in CONSTITUTIONAL REFORM IN THE UNITED KINGDOM: PRACTICE AND PRINCIPLES
85-97 (1998); Grant Huscroft & Paul Rishworth, “You Say You Want a Revolution:” Bills of Rights in
the Age of Human Rights, in DAVID DYZENHAUS, MURRAY HUNT, GRANT HUSCROFT EDS, A SIMPLE
COMMON LAWYER: ESSAYS IN HONOUR OF MICHAEL TAGGART (2009).
92
See Grant Huscroft, Civil Remedies for Breach of the Bill of Rights, in RISHWORTH ET AL, supra
note 51, at 837.
93
Andrew & Petra Butler, supra note 54, at 1116.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 22
My own more comparative evaluation is that, as in Canada, there is no
serious disagreement that, generally-speaking, rights are more effectively protected
under NZBORA than before. The original New Zealand system without a formal
declaration of incompatibility arguably provided insufficient signaling to the
legislature of a rights violation and thus not high enough political costs for ignoring
it, although the somewhat belated and partial legislative remedy following Quilter
may be seen as evidence both for and against this claim. On the other hand, without
the outlet – or buck-passing possibility – of the declaratory power at all, courts might
well feel greater pressure to employ the interpretive power in a strained or aggressive
way – and this may help to explain both the pre-Moonen more expansive line of
cases and the post-Moonen retreat from it, as exemplified by Hansen. With the
(implied) formal declaratory power, the somewhat narrower conception of the
interpretive duty makes sense to balance things out– but only if the power is used,
and so far the New Zealand courts have been reluctant to do so. Nonetheless, the
general assessment of NZBORA by its major commentators suggests something of a
corrective to the “no different than US-style judicial review” verdict on the new
model.
C. The United Kingdom
The central features of the HRA, which came into effect on October 2, 2000,
are well-known. They are: (1) a statutory bill of rights incorporating by reference
most, but not all, of the rights in the European Convention on Human Rights
(ECHR); (2) a duty on the responsible minister to make a statement as to whether or
not a proposed bill is compatible with the protected rights; (3) a requirement that
courts take the jurisprudence of the European Court of Human Rights (ECtHR) into
account in HRA cases; (4) a statement in section 6 that the rights binds courts and the
executive but not Parliament, which is also understood to mean that the courts can
directly invalidate executive action violating rights (unless that action is required by
statute) and are themselves bound to act in accordance with the rights, raising
complex and unresolved issues of the common law and horizontal effect; (5) an
obligation on the courts to interpret primary and secondary legislation consistently
with protected rights “so far as it is possible to do so” (section 3); (6) where not
possible, higher courts are empowered to issue a declaration of incompatibility (DI),
but this does not affect the continuing validity of the statute in the case itself or
generally (section 4); and (7) in response to a DI, the relevant minister may use a
special fast track legislative procedure to amend or repeal the statute--and, of course,
Parliament may do so through its normal legislative process.
The assessments of the HRA in practice by UK commentators do not exhibit
the broad consensus seen in New Zealand, but literally span the spectrum of possible
evaluations. Again, to the seemingly significant extent that they track pre-existing
attitudes towards HRA around the time of enactment, it is not clear how much of a
post hoc, evidence-based, assessment is being made. Here is the spectrum:
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 23
1. The “futility thesis” of Professor Keith Ewing. This contends that the HRA has
not had significant impact on or changed much in the UK’s legal and political system
and, in particular, has resulted in no better protection of rights.94
According to
Ewing, HRA has failed to prevent the enactment of rights-violating legislation and
judges have failed to oppose it. Only one piece of post-HRA Labour government
legislation has been the subject of a DI – section 23 of the Anti-Terrorism, Crime and
Security Act 2001 -- which is thus the exception not the rule. Otherwise, the
judiciary has exhibited traditional deference to the government and the pull of
parliamentary sovereignty remains strong.95
Ewing aside, there have been quite
general complaints of disappointing “judicial minimalism” in the interpretation and
application of the protected rights by the courts, especially in the criminal context, as
compared with the requirement of “generous” and “purposive” judicial rights
interpretation in Canada and New Zealand.96
Part of this criticism reflects a
widespread sense that UK courts are acting less independently in developing HRA
jurisprudence than expected and are too mechanically applying the jurisprudence of
ECtHR, even though they are not bound by it under HRA but only required to take it
into account.97
Adam Tomkins has argued that the courts’ poor record in protecting
rights should not be attributed solely or mainly to deference and the pull of
parliamentary sovereignty because judges have boldly asserted their new powers,
particularly over the executive, just not in a way that has protected civil liberties very
much.98
2. HRA is too weak. This is not because judges are failing to respect rights (as
mostly in the futility thesis), but because the government is – and the courts lack the
power to force it to.99
So, despite the judicial declaration in the Belmarsh Prisoners
Case that section 23 of the Anti-Terrorism, Crime and Security Act of 2001 --
authorizing indefinite detention of only foreign terror suspects -- was incompatible
with the defendants’ incorporated rights to liberty under Article 5 of the ECHR and
to non-discrimination on the ground of nationality under Article 14, the government
94
Keith Ewing, The Futility of the Human Rights Act, [2004] PUB. L. 829; Keith Ewing & J.
Tham, The Continuing Futility of the Human Rights Act, [2008] PUB.L. 668.
95
Not, however, that Ewing supports a stronger bill of rights or judicial review, as a well-known
opponent of both.
96
For example, Ian Leigh, Concluding Remarks, in Helen Fenwick, Gavin Phillipson, and Roger
Masterman, eds., JUDICIAL REASONING UNDER THE UK HUMAN RIGHTS ACT 443 (2007)
(“Minimalism is the dominant mode of (judicial) reasoning. An interim term report [of the judiciary’s
performance] would be along the lines of ‘tries hard, could do better.’)”. “Minimalism” was described
in an important judgment by Lord Bingham as the proper mode of interpretation, under which UK
courts should not expand the Convention rights beyond their interpretation by the ECtHR. See R. (on
the Application of Ullah) v. Special Administrator [2004] UKHL 26.
97
See Roger Masterman, Aspiration or Foundation? The Status of the Strasbourg Jurisprudence
and the ‘Convention Rights’in Domestic Law, in id.
98
Adam Tomkins, The Rule of Law in Blair’s Britain, 26 U. QUEENSLAND L.J. 255 (2007).
99
Robert Wintermute, The Human Rights Act’s First Five Years: Too Strong, Too Weak, or Just
Right?, 17 KING’S COLL. L.J. 209 (2006).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 24
did not choose to release the prisoners but waited several months until the new
control order legislation was enacted to “free” and immediately re-arrest them.
According to Robert Wintermute, the HRA record proves that the UK needs a
constitutionalized bill of rights and judicial power to invalidate conflicting
legislation, along the lines of the Canadian Charter.
3. HRA is too strong. It has in practice created de facto judicial supremacy so that
the difference between it and US-style judicial review is far less than anticipated or
hoped for. The evidence cited for this thesis is as follows:
(1) there has been no separate or independent political rights review to speak of, as
the pre-enactment review that has occurred has been dominated by legal discourse
and predictions of what the courts will ultimately do.100
(2) all but two surviving DIs have been dutifully complied with by the Labour
Government and in neither remaining case has it or Parliament announced plans to
exercise the legal power to ignore a declaration;101
(3) the courts have treated the section 3 interpretive duty as a very strong one and as
the primary technique of rights protection and remedy under the HRA. As a result,
courts have (a) not relied on section 4 – the main structural vehicle for legislative
voice—as much as expected or hoped, (b) effectively subjected Parliament to the
Convention rights despite the express wording of section 6; and (c) been rewriting
statutes at will. From this perspective, the future success of HRA depends on
whether there are effective ways to weaken judicial power, bolster political rights
review, and strengthen the willingness of parliament to challenge the courts.
4. A fourth view in essence agrees with this factual evaluation of the HRA but not
the normative. That is, the strength of the HRA, and in particular of the interpretive
obligation under section 3, is proper and justified, inherent not deviant, and creates
an alternative technique to judicial invalidation for protecting rights that is not only
its truly distinctive feature but also gives UK courts “powers of constitutional review
that are not hugely dissimilar from those possessed by the U.S. Supreme Court.” 102
Indeed, this view is skeptical that the HRA embodies “a distinctive form of
constitutionalism which can be contrasted so easily and so favourably with the U.S.
Constitution.”103
5. HRA is about right. A final group of academic commentators takes the view that
the HRA is working more or less successfully and as anticipated, with the courts
doing a satisfactory job of protecting rights and taking a reasonably balanced
100
Hiebert, supra note 2; Danny Nicol, The Human Rights Act and the Politicians, 24 LEG. STUD.
451 (2004).
101
See Hiebert, id; IAN LEIGH & ROGER MASTERMAN, MAKING RIGHTS REAL: THE HUMAN
RIGHTS ACT IN ITS FIRST DECADE 118 (2008).; Aileen Kavanagh, supra note 11, at 282-83.
102
Kavanagh, supra note 11, at 418
103
Id.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 25
approach to their interpretive and declaratory powers.104
This general position is also
the official one of the government, as expressed in a report published by the
Department of Constitutional Affairs in 2006.105
Less officially, however, several
government ministers and former ministers have disagreed with this claim, arguing
that HRA was a major mistake, unexpectedly handcuffing the government in the-then
unforeseen post 9/11 world.106
My own assessment falls somewhere between the third and fifth; that is,
between the view that HRA is too strong and that it is about right in practice.
On the futility thesis, I think this seriously underestimates the extent to which
courts have protected rights. Although it is true that only one DI has been made with
respect to post-HRA Labour legislation, (1) that one, the Belmarsh Prisoners’
Case,107
was of the highest importance and compares favorably with the fact that the
United States Supreme Court has not yet clearly and unequivocally held that
anyone’s constitutional rights have been violated by post 9/11 government action –
executive or legislative,108
and (2) 14 other DIs have been made and survived appeal.
Apart from the inevitable time-lag in the appeals process, some judicial caution in
choosing those cases in which to exercise a contentious new power is hardly
unprecedented or unexpected. Moreover, several other important cases have resulted
in judicial invalidation of executive action for violation of the HRA, including one of
the three House of Lords’ control order decisions.109
Finally, the section 3
interpretive power has also been an important judicial tool for the protection of rights
– particularly in such areas of sexual orientation discrimination110
and reverse onus
104
This group includes Alison Young, Conor Gearty and David Feldman.
105
Department of Constitutional Affairs, Review of the Implementation of the Human Rights Act
(2006).
106
The political future of HRA is currently highly uncertain. Absent any sort of public
consultation, it has never been particularly popular – many of the most visible cases have been
perceived as protecting terrorists and asylum seekers. The opposition Conservative Party has said it
will repeal HRA and replace it with a “British” bill of rights, but the content, status, and political will
for this remain unclear. Indeed, as at the time of writing, all three major political parties have
indicated that they would make changes to the HRA if elected at the upcoming general election. More
generally, the immediate political futures of both Canada’s section 33 and the NZBORA look a little
more stable. Despite the positive recommendation of the National Consultation Committee in
Australia, it is highly questionable whether the political will exists within the current Labor
government to introduce a national human rights act.
107
A and others v. Secretary of State for the Home Department [2004] UKHL 56.
108
Boumedienne v. Bush, 553 U.S. 723 (2008), and Hamdi v. Rumsfeld, 542 U.S. 507 (2004),
have come the closest. In Boumedienne, the Supreme Court held that aliens at Guantanamo Bay are
protected by the constitutional right to habeas corpus and so can raise constitutional claims
challenging their detention by the U.S. government. Cases raising the merits of those claims are now
making their way through the U.S. federal court system, with several detainees ordered released by
the lower courts. Although in Hamdi the Supreme Court held that in the case of a U.S. citizen, the
due process clause requires some opportunity to rebut the factual basis for the government’s
determination of enemy combatant status, only the dissenting justices clearly found a constitutional
violation in the case.
109
Secretary of State for the Home Department v. JJ and others [2007] UKHL 45.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 26
of criminal proof provisions111
-- and as already canvassed, it is hard to make the
case that this protective power has been meekly used.
On the general criticism of judicial minimalism, the large shadow of the
ECtHR over the UK system – disappointed litigants can still go to Strasbourg after
exhausting domestic remedies – has perhaps hampered the independence of both
courts and Parliament in their approaches to the protected rights. In this way, the
HRA is arguably a skewed example of the new model in practice. On the other hand,
given that UK courts are legally obligated only to take the ECtHR’s jurisprudence
into account, binding themselves to interpret Convention rights “no less” expansively
(even if also no more) than the international human rights court commonly regarded
as the most powerful and successful in the world is not to be sniffed at.
On the too weak thesis, to my mind, the major example cited of the Belmarsh
Prisoners’ Case illustrates the distinctive feature of the new model and how it is
meant to work rather than constituting a departure from expectation and a failure in
practice. At least under the HRA version of the new model as compared with the
Canadian, it was never intended that the judiciary would have the power to disapply
the incompatible legislation under which the government is acting and grant the
claimants redress – the legislature is legally free to express its disagreement with the
courts and do nothing.112
But in fact, the government and legislature did
affirmatively respond to the DI by subsequently amending the statute to create the
control order regime in place of executive detention in prison after prolonged and
very extensive parliamentary debate – and subject the prisoners to it. Indeed, in
counting this DI among all the others that have triggered such responses, it is part of
the evidence of those who argue that the HRA has proven too strong in practice.
Again, this failure of a judicial remedy under the DI mechanism was always part of
the distinctive HRA package. So while one might legitimately argue that, all things
considered, it is a flawed provision, it is not the case that practice under the HRA has
disclosed a problem that had not been initially identified or discussed. This is just an
intrinsic difference between the HRA and the Canadian versions of the new model.
On the too strong thesis, I think the critics have some reasonable grounds for
disappointment on this score given how the HRA has operated during its first decade,
but I think as an overall assessment it is significantly overstated and premature,
110
Ghaidan v. Godin-Mendoza [2004] UKHL 30.
111
R v. Lambert, supra note 54.
112
A section 3 interpretation of the statutory provision consistent with the Article 14 prohibition
of discrimination on grounds of national origin was not possible because the statutory provision
expressly applied only to foreign terror suspects. Moreover, given the DI, judicial invalidation of the
executive orders for indefinite detention was not possible, under section 6 (2) (b) of the HRA. This
provision states:
6 (1). It is unlawful for a public authority to act in a way which is incompatible with a
Convention right.
(2) Subsection (1) does not apply to an act if―
… (b) In the case of one or more provisions of…primary legislation which cannot be read or
given effect in a way which is compatible with the Convention rights, the authority was acting so as to
give effect to or enforce those provisions.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 27
especially to the extent it argues there is little difference in practice between the HRA
and de jure judicial supremacy.
Those hoping for specifically political pre-enactment rights review, as a way
of bringing a different type of rights-consciousness to bear on the issues, have
perhaps been most disappointed in that there has been far too much focus on legal
argument and predicting what courts (both UK and ECtHR) will do. On the other
hand, the work of the Joint Committee on Human Rights (JCHR) as a high-quality,
cross-party parliamentary committee has been quite successful in alerting Parliament
to rights concerns.113
Indeed, the JCHR has criticized the courts for not protecting
rights aggressively enough, so as a parliamentary body it seems not to think there is
de facto judicial supremacy. Also, the fact that under HRA, the responsible minister
must make an HRA statement whether or not she deems the bill incompatible is
preferable to the situation in Canada and NZ where such statements are only required
if the bill is deemed incompatible. Despite only one incompatibility statement so far,
a significant amount of HRA-conscious work is undertaken by the government and
its civil servants prior to a bill’s introduction, even if too much of this is currently
legal and predictive in nature. Overall, and contrary to the general thrust of this
source of criticism, there is good evidence that the HRA is making some difference
in the legislative process – although there is undoubtedly room for more.114
On the over-compliance with DIs, the political branches are between a rock
and a hard place. So on the one hand, to the extent that the greatest substantive
concern with the new model as a whole is inadequate protection of rights without a
constitutionalized charter and a non-overridable judicial invalidation power, routine
ignoring of DIs would undoubtedly be taken by some as clear evidence that HRA is
too weak and that the model is unstable, reverting to traditional parliamentary
sovereignty.115
Yet, on the other hand, the record of compliance then prompts the
claim that the legislature and executive are just kowtowing to courts, afraid to
exercise an independent voice, so that the model is unstable in the other direction,
rolling towards judicial supremacy.
As I outlined earlier under “ideals,” I think the best understanding and
operation of legislative final words in general is a presumption that legislatures will
abide by court decisions and not routinely ignore them – but on matters of principle
where there is reasonable disagreement after serious debate, it should be considered
legitimate for legislatures to exercise their independent legal power of the final word.
The deliberative process is, however, more important than the outcome per se. As
perhaps also in Canada, it is not entirely clear that such issues have really arisen yet.
Even the Belmarsh Prisoners’ Case was decided on, and limited to, the largely
113
Andrew & Petra Butler, supra note 54, at 198 n3; Janet Hiebert, Parliament and the Human
Rights Act: Can the JCHR Help Facilitate a Culture of Rights, 4 ICON 1 (2006).
114
See, e.g., Hiebert, id; Francesca Klug & H. Wildbore, Breaking New Ground: the Joint
Committee on Human Rights and the Role of Parliament in Human Rights Compliance, EUROPEAN
HUMAN RIGHTS LAW REVIEW 231 (2007).
115
Of course, Ewing makes this argument about reversion as part of his futility thesis even
though Parliament has complied, or is in the process of complying, with all surviving declarations of
incompatibility.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 28
indefensible issue of arbitrary (i.e., unreasonable) discrimination against non-UK
citizens rather than the more difficult and contestable underlying issue of the
permissibility of indefinite detention itself. Although Parliament did not ignore the
DI but responded to it after several months, and in that response fully accepted the
discrimination point by making the new control order regime applicable regardless of
nationality, this new regime, adopted after one of the longest Parliamentary debates
in British history, reflected something of a pre-emptive compromise on the
underlying issue – a compromise that was subsequently litigated in the “second
look” control order cases.116
In addition, there is evidence of significant legislative
consideration and debate in response to several of the other DIs.117
Overall, perhaps
this is as it should be.
On the alleged judicial overreliance on section 3, I think at the outset it is a
mistake categorically to view every use of section 3 rather than section 4 as a failure
of institutional balance or to conceive of section 4 as the exclusive basis for
legislative voice. First, as enacted, section 3 is undoubtedly a central part of the
overall balance between rights protection and democratic decision-making under the
HRA, part of its complex institutional design and distribution of powers, and not
merely an afterthought or stepping stone to section 4. It is always the first step in
any case and by the logic of the section 3/section 4 divide, always likely to be used
in many cases – where it is at least “reasonably possible” to give the statute a
meaning consistent with the relevant right. Certainly, although hardly conclusive,
the Government argued during enactment of HRA that most rights claims would be
resolved by section 3 and resort to section 4 would rarely be necessary.
Second, use of section 3 does not deprive Parliament of the final word, at
least not to the extent the critics seem to suggest. As Lord Steyn properly made clear
in Ghaidan, Parliament can always respond to a section 3 decision by amending the
legislation to overrule the judicial interpretation and make clear its view that the
courts have mistaken its meaning or overstepped their authority – subject of course
to the possibility of a subsequent DI and/or a likely losing case at the ECtHR.118
As
someone who teaches U.S. constitutional law, it is strange to hear that the use of
section 3 deprives Parliament of the final word, when the traditional distinction
between statutory and constitutional interpretation under judicial supremacy is
precisely that under the former but not the latter, legislatures remain free to – and
sometimes do in practice -- overrule the court’s interpretation by amending the
legislation.119
It is, of course, true (a) that section 3, unlike section 4, places the
116
See supra note 79. These cases perhaps suggest a more valid criticism of the courts as being
more comfortable reviewing executive action rather than the underlying legislation for HRA
compatibility, see Ewing, The Continuing Futility of the Human Rights Act, supra note 90. Again, this
may be due not to parliamentary sovereignty but to the fact that this way, unlike under Section 4, they
can grant a remedy to the complainant.
117
See Alison L. Young, The Practicalities of Dialogue: Is Dialogue Working under the Human
Rights Act 1998? (manuscript on file with author).
118
“If Parliament disagrees with an interpretation by the courts under Section 3(1), it is free to
override it by amending the legislation and expressly reinstating the incompatibility.” Ghaidan, supra
note 106, at para. 63.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 29
default position in favor of the courts and (b) that it does not involve “ordinary”
statutory interpretation. I do not believe, however, that the change from ordinary to
presumptively rights-consistent statutory interpretation is either so novel/distinctive a
technique comparatively,120
or that it so transforms the courts’ task as to make a
legislative response practically impossible in all political contexts. Would it
necessarily be politically impossible, say, for a newly elected Conservative
Government to change its mind about repealing the HRA but to amend the Rent Act
to overrule Ghaidan instead, if it believed that popular sentiment supported it?
Finally on the way section 3 has been used, I think it also overstates the case
to claim that the practice of the courts have been so aggressive and radical as to
amount to rewriting legislation at will, making rather than interpreting statutory law–
thereby violating sections 3(b)(2),121
4 and 6 by failing to give effect to incompatible
statutes. It is certainly true that, overall, the UK courts have treated the interpretive
obligation as a strong one, a little stronger than New Zealand courts, and have
advanced beyond pre-existing domestic modes of statutory interpretation – even the
most advanced common law fundamental rights protective ones developed in the
years immediately before the HRA. On this latter point, it is only necessary to
compare Fitzpatrick, a 2001 pre-HRA case, with Ghaidan, a 2003 post-HRA case,
interpreting the very same provision of the 1988 Rent Act Amendment in opposite
ways.122
Indeed, it is probably accurate and candid to describe the judicial function
under section 3 as including limited modifications of statutes to protect rights rather
than merely interpretation.123
At the same time, however, (1) as in New Zealand,
different judges take different views of the appropriate strength of the interpretive
obligation so, as one recent commentator puts it, “[t]o distil or elicit any working
principles or rules of guidance from the cases is not easy”124
; (2) there are still
undoubtedly limits – including inherent ones -- beyond which a Convention-
119
For example, Congress has over the last couple of years been in the process of overriding
several restrictive, pro-employer interpretations of the 1964 Civil Rights Act and other antidiscrimination
statutes.
120
See infra note 130.
121
“This section [Section 3] does not affect the validity, continuing operation or enforcement of
any incompatible primary legislation.” HRA section 3(2)(b). On this textual constraint on the scope
of the interpretive obligation under section 3, see Conor Gearty, Reconciling Parliamentary
Democracy and Human Rights, 118 L.Q.R. 249 (2002).
122
Fitzpatrick v. Sterling Housing Association Ltd [2001] 1 A.C. 27. For a detailed discussion of
the differences between the two cases, see Kavanagh, supra note 11, at 108-114. More generally,
however, Kavanagh argues that the courts do not have radically new interpretive powers under the
HRA; the difference is primarily that they have used their powers more frequently and with a greater
sense of legitimacy.
123
See Kavanagh, supra note 11, at 99.
124
Alec Samuels, Human Rights Act 1998 Section 3: A New Dimension to Statutory
Interpretation?, 29 STATUTE L. REV. 130, 138 (2008).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 30
compatible interpretation is not possible,125
and (3) in total, the courts have only used
section 3 to “rectify statutory language in a small number of cases.”126
Moreover, a good argument can be made that the general principles for
applying section 3 have slightly weakened over time from the 2002 high point of
interpretive power in R v. A.127
Whereas in that case, the leading opinion of Lord
Steyn stated that a rights-consistent meaning is not “possible” if -- and perhaps only
if -- Parliament has clearly and expressly limited a convention right,128
in Ghaidan,
decided in 2004, several of their Lordships clarified that a rights-consistent
interpretation is not possible where it would be incompatible with “the underlying
thrust of the legislation” (Lord Nicholls),129
or “the fundamental features of the
legislative scheme” (Lord Millett),130
or would not “go with the grain” of the
legislation (Lord Rodger).131
And arguably, Wilkinson, decided the following year,
rolled back the standard of impossibility a little further still.132
Here, Lord
Hoffman’s leading opinion rejected the notion, arguably implied in R v A, that
section 3 required courts to give the language of statutes “acontextual meanings,”133
and stated that “the question is still one of interpretation, i.e., the ascertainment of
what, taking into account the presumption [that Parliament did not intend a statute to
mean something which would be incompatible with those rights] created by section
3, Parliament would reasonably be understood to have meant by using the actual
language of the statute.”134
Indeed, if one looks at the application of the stated section 3 principles in all
three cases, the impression of a slight weakening over time is further confirmed.
Thus, it seems hard to conclude that the section 3 interpretation of the rape shield
law at issue in R. v. A – reading in a fair trial limitation -- did not fly in the face of
Parliament’s specific intent or the “underlying thrust” of the statute as manifested in
its unambiguous blanket exclusion of evidence of prior sexual history between the
defendant and complainant. By contrast, in Ghaidan, it does not seem an
unreasonable contextual reading of a statute extending tenure protection, absent other
clear language to the contrary, to interpret the words “living together as his or her
wife or husband” to refer to a general relationship of sexual intimacy exemplified by
but not limited to the heterosexual relationship of husband and wife. Indeed, Lord
125
On limits, see infra text accompanying notes 122-127.
126
Kavanagh, supra note 11, at 117.
127
R. v. A (No 2) [2002] 1 AC 45.
128
Id.. at para. 44 (Steyn LJ).
129
Ghaidan, supra note 104,at para 33.
130
Id., at 67.
131
Id., at paras. 110, 111, 116.
132
R. v. Her Majesty’s Commissioners of Inland Revenue ex parte Wilkinson [2005] UKHL 30.
133
Id., at para. 17.
134
Id (emphasis in original).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 31
Nicholls argued that the “social policy” underlying the statutory extension of tenure
to the survivor of couples living together as husband and wife “is equally applicable”
to the survivor of homosexual couple living together in a close and stable
relationship.135
And in Wilkinson, because there were such clear contrary indications
elsewhere in the statute, the House of Lords found that no “reasonable reader could
understand the word ‘widow’ to refer to the more general concept of a surviving
spouse,”136
so that a compatible, non-discriminatory interpretation of a tax provision
granting a bereavement allowance only to widows was not possible.
What about the fourth view, that the strength of the HRA is not only factually
established but normatively justified, that section 3 is the truly distinctive judicial
technique, and that overall UK courts have powers not “hugely dissimilar” from
those in the U.S., thereby challenging the claim that the HRA embodies a new or
distinctive model of constitutionalism at all? I will say more about the issue of the
general distinctiveness of the new model in the next section. But specifically on the
HRA, and bracketing the normative claim for the moment to focus on the
descriptive, I believe – as just argued – that this overstates the strength of judicial
power under the HRA and also understates its distinctiveness. On the latter, I
continue to think that section 4 (and its subsequent equivalents in Australia and New
Zealand) is the novel and distinctive – indeed unique -- judicial rights-protecting
technique, which gives defining shape to the HRA overall,137
more so than section 3,
which has fairly common pre-existing equivalents among various types of
contemporary legal regimes, including the EU, NZBORA, and the German Basic
Law.138
The fact that Parliament has not so far used its legal power of the final word
is not (yet) compelling evidence that it is practically irrelevant; it is far too early to
draw this conclusion.
Moreover, in conceptualizing the HRA (and the new Commonwealth model
of constitutionalism) as an intermediate or hybrid form of constitutionalism, it is far
from clear that “hugely dissimilar powers” are claimed or to be expected. As a
matter of general comparative constitutional methodology, it perhaps smoothes over
135
Ghaidan, supra note 104, at para 35.
136
Wilkinson, supra note 132.
137
At the time of HRA’s enactment, no other system of constitutional review of legislation in the
world – domestic or international, past or present, contained the same or a similar judicial power. As
we’ve seen, this power was subsequently been implied by the New Zealand Court of Appeal, although
so far not clearly used. It was also enacted as part of both the ACT HRA and the VCHRR. The
SCC’s suspended declaration of invalidity is quite different in that the legislature acts in the shadow of
a legally authoritative reversion to a judicial order invalidating the relevant statute.
138
In the EU, the Marleasing case imposed a duty on all national courts to interpret domestic law
in line with untransposed directives wherever possible to do so; Marleasing SA v. La Comercial
Internacionale de Alimentacion SA, Case C-106/89 [1990]. In New Zealand, section 6 of the New
Zealand Bill of Rights Act, 1990 states that: “[w]herever an enactment can be given a meaning that is
consistent with the rights and freedoms contained in this Bill of Rights, that meaning shall be
preferred to any other meaning.” In Germany, the famous Lüth decision of 1958 stated that:
“Constitutional rights form an objective order of values….This value system…must be looked upon
as a fundamental constitutional decision affecting the entire legal system….It naturally influences
private law as well; no rule of law may conflict with it, and all such rules must be construed in
accordance with its spirit.” Lüth, BverfGE 7, 198 (1958) (emphases added).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 32
too much -- or is too “similarity-oriented”139
-- to deny that lesser differences are
significant or relevant. Indeed, the three models of constitutionalism are most
essentially about different allocations of legal powers among institutions, so while it
is of course appropriate and valuable to ask how they interact with other factors in
practice, such as legal culture and political reality, this is a perennial and distinct
issue. Even before the HRA, of course, it was sometimes claimed that protection of
rights in the UK was in practice not that much different from the US despite the clear
– and, in a non-trivial sense, irreducibly important -- legal differences in respective
constitutional systems.
Focusing briefly on the other side of the equation, it seems hard to believe
that replacement of judicial supremacy by an HRA-style regime would make no
practical difference in countries such as the United States and Germany. Who would
wager that if subject only to a declaration of incompatibility or a modifying, rightsconsistent
interpretation, statutes such as the one at issue in Roe v. Wade140
would not
have been saved by exercise of a legislative final word, or the Bavarian school law
litigated in the FCC’s second crucifix case?141
Would citizens likely be persuaded
that the change in rights regime is purely formal only? Moreover, Ireland did enact
an HRA-like statute for its domestic incorporation of the ECHR in 2003, even
though it has long had a constitutional bill of rights and a judicial power of
invalidation to enforce it.142
Why would Ireland do this unless it was thought to
make some practical difference? The answer cannot simply be that it was easier
politically to enact a statute than a constitutional amendment because Ireland could
still have incorporated a judicial power to invalidate legislation inconsistent with the
ECHR into the statute.
Overall, with respect to the first criterion of success, it is I think hard to
establish that there is not now more effective rights protection than before HRA.
Admittedly, the radical change in political context and national security exigencies
brought about by 9/11 has complicated the direct comparison of pre-and post-HRA
regimes. Moreover, it is very difficult to assess the complex counterfactual of what
purely legislative rights protection absent the HRA would now look like.
Nonetheless, it seems to me hard to deny that rights are better protected under the
HRA in the following specific ways. There is now greater rights consciousness than
before -- among citizens, courts, Parliament and government -- and the rights that
exist are generally better and more widely known and understood than under the preHRA
regime of common law rights supplemented by various specific statutory
protections. There are also more legally recognized rights. Pre-HRA, there would
likely have been no plausible legal claim for the UK courts to consider in the
Belmarsh Prison case as there was no obvious relevant domestic right against
139
Pierre Legrand, The Same and the Different, in PIERRE LEGRAND AND RODNEY MAUNDAY EDS,
COMPARATIVE LEGAL STUDIES: TRADITIONS AND TRANSITIONS (2003).
140
410 U.S. 113 (1973).
141
German Classroom Crucifix Case II, 93 BverfGE 1 (1995).
142
European Convention on Human Rights Act, 2003 (Ireland).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 33
statutory discrimination on the basis of citizenship, any more than there was on the
basis of sexual orientation in the pre-HRA case of Fitzpatrick.143
Finally, even
though for reasons discussed in the next section, the current system of remedies for
rights violations under the HRA is somewhat partial, unsatisfactory and in need of
reform, it is at the very least the same as – and so no worse than – the pre-HRA
regime, and arguably stronger.
On the second criterion, the balance of power has I think at most gone a little
too far in the judicial direction, taking into account the pre-enactment focus on what
the courts will do as well, and somewhat greater use of section 4 should certainly be
encouraged – perhaps by instituting the norm of a legislative remedy that I discuss in
the following section -- as should greater political rights review both before and after
the judicial stage. But only a little. Moreover, to the extent that the two criteria
themselves must be balanced, I think a thumb on the scale of adequate rights
protection is justified. All in all, my view is that HRA is working reasonably well
and that there is no clear and obvious practical problem with this version of the new
model, as there is for example in Canada.
V. FUTURE DIRECTIONS FOR THE NEW COMMONWEALTH MODEL
As we have seen, academic commentators in the individual countries have
provided mixed reviews of the various versions of the new model in practice,
especially in Canada and the UK. On the two general criteria of (1) are rights
adequately protected and (2) are there appropriate limits on, and a better balance
between, legislative and judicial power, it is fair to say that (with obvious
exceptions) most of the expressed concerns have been less with (1) than (2).144
And
here, probably the main criticism is that developments in practice such as a culture of
legislative compliance with courts and overreliance on interpretive techniques for
protecting rights have caused the balance to veer too much in favor of judicial power
– so diminishing or undermining the distinctiveness of the new model.
In theory and form, it seems clear that the new Commonwealth model is a
distinctive, intermediate model of constitutionalism. In granting the final legal word
to the legislature, both Canada’s section 33 and the UK’s declaration of
incompatibility were novel provisions when enacted and are alien to and inconsistent
with the standard modern system of constitutional review. The legal requirement of
pre-enactment rights review by the political branches is also an important and
attractive part of what makes the model new. But even now, as practiced, while
acknowledging that some of the critics’ concerns are justified, the new model
remains more distinctive than this predominant line of criticism claims.
143
Fitzpatrick v Sterling Housing Association Ltd [2001] 1 A.C. 27.
144
Indeed, Ian Leigh and Roger Masterman have argued that in the UK context there has been
“neglect” of the first (“and most important”) criterion of success in favor of the second. IAN LEIGH
AND ROGER MASTERMAN, MAKING RIGHTS REAL: THE HUMAN RIGHTS ACT IN ITS FIRST DECADE 235
(2008).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 34
Generalizing from what I have said above and attempting to place an overall
assessment in proper comparative constitutional perspective, it is necessary to bear in
mind three key features of the new model. First, as long as it is not truly politically
unusable, bestowing the final legal word on the legislature is still the critical
difference between the new model and de jure judicial supremacy. This is because it
is still the case that a truly controversial, unpopular court decision may very likely be
overridden, even in Canada, and this in turn may help to explain why so few, if any,
such court decisions have been rendered. There is as a result a more specific,
immediate, and direct public opinion constraint on courts than in such paradigmatic
strong jurisdictions as the US and Germany, where the only options are passing a
constitutional amendment, attempting to secure future judicial appointees with
different views, or taking to the streets. Second, as institutionalized in previously
Westminster-based systems, the new model obviously does and is intended to
increase judicial power compared with traditional parliamentary sovereignty and
should be candidly acknowledged as such. The relevant question, accordingly, is not
whether there has been an increase in judicial power but whether the right amount,
too much, or too little. Third, as important a criterion of success as institutional
balance is, it is not the sole one and may itself have to be balanced against the other:
greater and effective protection of fundamental rights. Assessments that do not take
this into account and focus only on legislative-judicial power are, I believe,
overlooking an essential factor.
Nonetheless, at least for those who find the new model normatively appealing
as an intermediate model of constitutionalism, the question arises of how might its
distinctive, intermediate features be bolstered in practice? One principal practical
dilemma has emerged from experience under the new model as a whole. If courts
are granted the power to invalidate rights-incompatible statutes (as under the
Canadian Charter), there is a serious risk that the political costs will generally be too
high for the legislature to exercise its override power – leading to more effective
rights protection at the cost of a de facto judicial final word. If, on the other hand,
courts are not granted the power to invalidate rights-incompatible statutes (as in the
UK, New Zealand and under the ACT HRA and the VCHRR), then the risk that the
individual whose rights they adjudicate to have been violated will not receive a
remedy creates strong pressure on the courts to rely too much on their interpretive
power/duty, again at the potential cost of ousting legislative judgment and creating a
de facto judicial final word. In each case, the risk is that the legal final word is in
practice made too hard to exercise where the default rests with the court decision, as
under both the invalidation and interpretive powers.
Although New Zealand courts have seemed better able than those of the UK
to resist the pressure under the second horn of this dilemma to overly rely on the
interpretive duty, and so as a result may be the more representative standard-bearer
of the model than the ECtHR-skewed HRA, the general solution I believe lies in
reducing the pressure. Returning to the section 3 versus section 4 issue under the
HRA, it is generally true, as commentators have suggested, that – rather than
necessarily the result of a bare desire to aggrandize themselves at the expense of
Parliament -- the remedial structure of section 3 and 4 creates strong pressure on
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 35
courts to use the interpretive power under which they can grant a remedy to the
aggrieved individual versus section 4, under which they cannot.145
Slightly less
noted is the fact that the same remedial pressures also push the courts into assessing
specific executive action, which they generally have the power to invalidate,146
rather
than the underlying statutes―as, for example, in the control order cases. Although
as I have said, I do not think use of section 3 is quite as much of a problem from the
perspective of legislative voice as critics suggest, the political branches have not
done all they can to alleviate the problem and so reduce this pressure on the courts.
Under the “fast-track” procedure of HRA section 10, there is express
provision for the amending order to have retrospective effect and therefore to grant a
legislative remedy to the individual whose rights the court has declared have been
violated.147
However, this power has not been used,148
primarily because all but one
of the surviving DIs have been responded to by amending legislation through the
ordinary parliamentary process rather than section 10 (a procedural record perhaps
otherwise to be approved of). So where the government intends to comply with a DI,
it could and should either (1) use the express section 10 power to make orders
retroactive, or (2) ensure that, wherever possible, along with the amending
legislation – whether or not it is given retroactive effect -- provision is made to
afford a remedy to the individuals affected by the incompatibility.149
Indeed, Ireland
has expressly included a discretionary governmental power to award an ex gratia
payment of compensation to the successful litigant following a DI under its statute
incorporating the ECHR into domestic law.150
Moreover, the duty to repair the
damage caused to individuals and to return them to their ex ante position is already a
legal obligation of the UK under the ECHR whenever the ECtHR finds a
145
Aileen Kavanagh, Choosing between Sections 3 and 4 of the Human Rights Act 1998: Judicial
Reasoning after Ghaidan v, Mendoza, in HELEN FENWICK, GAVIN PHILLIPSON, ROGER MASTERMAN
EDS., JUDICIAL REASONING UNDER THE HUMAN RIGHTS ACT (2007) (remedial concerns have been
central to the courts’ use of Section 3 versus Section 4). Rosalind Dixon, A Minimalist Charter of
Rights for Australia: The UK or Canada as a Model? 37 FEDERAL L. REV. 335 (2009).
146
For the exception under HRA, see supra note 108.
147
HRA, Schedule 2, Section 10, para. 1(1) (b): “A remedial order may be made so as to have
effect from a date earlier than that on which it is made.”
148
So far, “no amendments [to legislation in response to a DI] have been given retrospective
effect so as to afford rights to the individual at whose suit the declaration was obtained.” See JACK
BEATSON, STEPHEN GROSZ, TOM HICKMAN, RABINDER SINGH, HUMAN RIGHTS: JUDICIAL PROTECTION
IN THE UNITED KINGDOM 37 (2008).
149
This has also been proposed by the JCHR as the last of five recommended steps the
Government should take to try and persuade the ECtHR that the DI is or has become an effective
remedy. See JCHR, Sixteenth Report of Session 2006-07, Monitoring the Government’s Response to
Court Judgments Finding Breaches of Human Rights, HL 128, HC 728, para. 119 (“[These steps
should include] ensuring that any legislative solution makes the necessary provision to afford a
remedy to the applicants affected by the identified incompatibility.”)
150
European Convention on Human Rights Act 2003 (Ireland), §5(4) (c). In addition, Canadian
legislatures have occasionally given legislation responding to Supreme Court of Canada decisions
retroactive effect. For an article discussing this phenomenon in Canada and advocating that the courts
adopt a general interpretive presumption that such legislation have retroactive effect, see Sujit
Choudhry and Kent Roach, “Putting the Past Behind Us? Prospective Judicial and Legislative
Constitutional Remedies,” 21 SUPREME COURT LAW REVIEW (2d) 205 (2003).
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 36
violation.151
It is also the typical remedy, in the form of “just satisfaction,” afforded
by the ECtHR itself against the member-state to the injured party where national law
permits only partial reparation to be made, under Article 41 of the ECHR.
Not only would the development of such a rule or norm potentially solve this
practical dilemma but it would also be fair. From the perspective of the new model
as a whole, I don’t think the absence of a judicial remedy after exercise of the
declaratory power is inherently unjust or problematic.152
If a court’s view of the
constitutionality of a statute is not the final word, I don’t think a remedy needs to be
granted at this point – but it does need to be granted after the final decision if that
decision is in favor of the party. Accordingly, where the legislature responds to a
declaration of incompatibility by amending or repealing the relevant statutory
provision and so acknowledges the problem, it should also, as part of that new
legislation, provide as much of a legislative remedy as it can, including either or both
monetary compensation and retroactive effect.
I do not see any basis for thinking there might be a legal problem for a
legislative remedy of compensating individuals affected by a law following a DI. As
for the separate issue of retroactive remedial orders or amending legislation, neither
do I believe there is likely to be any general bar to this second or supplementary type
of legislative remedy under section 4.153
First, the traditional common law
presumption of the non-retroactivity of statutes in the name of rule of law principles
is just that: a presumption that has always been deemed rebuttable by clear
legislative wording. Second, there is nothing in the HRA itself to suggest that
retroactive remedial legislation would violate any of the convention rights; to the
contrary, as just mentioned, it grants express power to the minister to introduce
remedial orders with retroactive effect under section 10.154
Third, the identical issue
of retroactivity arises under sections 3 and 4. If common law rule of law principles
(or any others) do not prevent changed and retroactive statutory interpretation
altering “vested” private rights under section 3 – as, for example, in Ghaidan – it is
unclear how or why they could prevent precisely the same thing via retroactive
legislative amendment. The section 3 decision in Ghaidan meant, of course, that in
2004 the landlord was bound by an interpretation of the 1977 Rent Act that differed
151
See JCHR Sixteenth Report of Session 2006-07, supra note 105 at 10, citing the relevant
ECtHR judgments to this effect. Indeed, such legislative compensation may thus also be a way of
pre-empting a subsequent successful appeal to the ECtHR following a DI.
152
Whether in the HRA context it is problematic from the perspective of the ECHR right to an
effective remedy, left unincorporated by the HRA, is a different issue and one the ECtHR may soon
answer and so preempt.
153
For a different view, see Rosalind Dixon, supra note 141. Neither of two recent works
focusing specifically on retroactivity addresses this precise point (as distinct from the issue of to
which events and causes of action pre-dating October 2, 2000 does the HRA apply): BEN
JURATOWITCH, RETROACTIVITY AND THE COMMON LAW (2008); David Mead, Rights, Relationships
and Retrospectivity: The Impact of Convention Rights on Pre-Existing Private Relationships
Following Wilson and Ghaidan, [2005] PUB. L. 459.
154
There is some language in HRA relevant to the entirely different issue of whether and when
HRA applies to events occurring or causes of action brought before it came into effect on October 1,
2009. On this issue, and not without a good deal of controversy, the courts have so far generally held
that HRA does not apply.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 37
from the one on which he could reasonably have relied when entering into the lease
with Godin-Mendoza’s same-sex partner in 1983. Fourth, the ECtHR has held that a
declaration of incompatibility under section 4 does not constitute an “effective
remedy” for the purposes of the ECHR procedural requirement of exhausting
domestic remedies155
– and so might also find that it does not satisfy the ECHR right
to an effective remedy before a national authority under Article 13, a right
deliberately omitted from those incorporated under the HRA. Accordingly, action to
strengthen this “weak remedy” along either of the suggested lines is, I think, unlikely
to face general obstacles from this source.156
Finally, another member of the House
of Lords Judicial Committee has opined that there is no such bar. In the course of his
dissenting speech in Ghaidan arguing that section 4 should have been used because
the non-discriminatory interpretation of the Rent Act was not “possible,” Lord
Millett stated that: “It [incompatible legislation] continues in full force and effect
unless and until it is repealed or amended by Parliament, which can decide whether
to change the law and if so from what date and whether retrospectively or not.”157
As far as practical problems are concerned, one or other – or some
combination -- of the two legislative remedies would seem possible and reasonably
effective in all types of cases. In criminal law, compensation can be paid for costs
suffered in the time between the judicial finding of incompatibility and the legislative
response. Even in a case like Ghaidan, although practically the losing tenant would
either have been evicted from the flat or paid more rent under a section 4 treatment
of the case, there is no reason why the government could and should not compensate
such tenants for their pecuniary loss resulting from the existence of a law on the
statute books that both it and the courts agree violates their rights.158
As mentioned
above, this is the approach that Ireland has taken in its statute domesticating the
ECHR. Moreover, the possibility of ultimately obtaining legislative or executive
compensation may help to counter any “plaintiff-disincentivising”159
effects of the
declaratory mechanism, although the facts that (1) there is generally ex ante
uncertainty as to whether judges will currently rely on it or the interpretive power in
any given case and (2) plaintiffs can seek both remedies in the alternative suggest
that these effects are small. Accordingly, the development of such a norm (or legal
rule) should significantly reduce any artificial distortion in the interplay between the
interpretive duty and the declaratory power and bring them into better balance.
155
Burden and Burden v. United Kingdom, 44 EHRR 51 (2007).
156
Even if, contrary to what I have argued above, the ECtHR were to hold that the ECHR
prohibits such legislative or executive remedies, this obviously wouldn’t apply to New Zealand or
Australia, or to the model generally.
157
Ghaidan, supra note 104, at para. 64.
158
This would have been the remedy afforded by the ECtHR in Karner v. Austria, an essentially
identical case to Ghaidan decided a year earlier, had the complainant survived or left beneficiaries of
his estate. (2003) 2. F.L.R. 623.
159
Andrew & Petra Butler, supra note 58, at 1116.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 38
To return to the ideal working of new model – or, rather, how we would
actually want it to work – discussed in Part III, what other changes in practice or law
does experience suggest might push the model further in this direction?
To enhance the first stage of pre-enactment rights review, experience suggests
that the relevant minister (and so in practice the department’s civil servants) should
always be required to make a statement about the compatibility of any proposed
statute with the bill of rights and not only when the minister deems the bill
incompatible (as currently in Canada and New Zealand). Moreover, the expectation
should be of serious ministerial consideration, not merely formulaic or conclusory.
Second, there must be focused parliamentary consideration of this statement and the
material on which it was based, as well as of the bill’s general compatibility with the
protected rights. Again, the UK experience of a specialized, non-partisan, and wellstaffed
parliamentary committee issuing a report on each bill, while not perfect,
suggests the best way to achieve this part of pre-enactment rights review. The more
this is conducted by political actors themselves rather than government and
parliamentary lawyers, the more likely that a specifically political rights review will
emerge – which in turn helps to justify a greater role for the elected branches relative
to the legal expertise of the courts.
Recall the two suggested ideal features of the model’s judicial stage as
compared with the typical workings of traditional strong review: (1) serious and
respectful consideration of the rights review undertaken at stage one by the elected
branches but, at the same time, (2) the duty to provide the political community with
an independent and high-quality legal judgment on the merits that not having
responsibility for the final word both requires and makes more possible. On the first,
obviously the relatively disappointing nature of political rights review thus far as
predicting what the courts will do has not created much space for such consideration.
On the second, again to the extent that the remedy dilemma has artificially distorted
the courts’ judgments, resolving it should result in less constrained, more
independent judicial decisions on using the declaratory versus interpretive powers.
To the extent one has faith in textual constraints, statutory provisions such as the two
Australian ones – enacted several years after the HRA -- requiring courts to exercise
their new interpretive duties “consistently with their [i.e., the legislative] purpose”160
may also have some impact.
At the third stage of legislative reconsideration of statutes deemed rightsincompatible
by the courts (or given strained, rights-compatible interpretations),
experience under the various versions of the new model suggests two things are of
potential importance. The first is the development of a norm of legislative remedy
discussed above. The second is that the conception or framing of the legislative
power of the final word should generally have the connotation of overriding a judicial
interpretation or decision on rights and not overriding the rights themselves – it
should in Waldron’s terms be understood to be about “rights disagreements” and not
“rights misgivings.”161
This is clearly not the case currently with section 33 in
160
ACT HRA, section 30; VCHRR, section 31(1).
161
See Waldron, supra note 19; Hiebert, supra note 22.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 39
Canada, and is also likely not so with legislative responses to declarations of
incompatibility in the UK. Such a conception would reduce the current high political
cost of exercising the legislative final word. If understandings cannot now be easily
changed, then textual amendments (or modified judicial statements in New Zealand)
may be necessary. Here, as the Evanses and Roger Masterman have suggested, the
Victorian Charter’s formula of a judicial declaration of “inconsistent interpretation”
may well be an improvement.162
A more radical alteration to the existing statutory versions of the new model
at the judicial stage, one that clearly weakens judicial power, would be to omit the
interpretive duty altogether (or seriously reduce it), and so authorize only “ordinary”
modes of statutory interpretation to protect the rights. This should result in greater
reliance on judicial declarations of incompatibility, which may in turn reduce the
political costs associated with ignoring them. On the other hand, of course, it may
result in perceived or actual underprotection of rights, depending in part on how
legislatures respond to declarations. This more radical change could be used either
instead of or in addition to the legislative remedy reform just discussed --- thereby
reducing both the power and the incentives of the judiciary to rely on it.
An even more radical alteration would be abolition of both the interpretive
duty and the judicial declaratory power (which, as Rishworth has shown, are
inherently connected) and to have simply a statutory bill of rights – at least as far as
legislation is concerned. Judicial power could be maintained for executive action.
Although certainly a possible option within a system of parliamentary sovereignty,
such a “pure” statutory bill of rights, however, would not be an example of the new
model, seemingly lacking as it does the second requirement of enhanced formal
judicial power concerning rights. A less radical alternative, at least formally, than
either of these two would be to revert to the original New Zealand model of a purely
interpretive bill of rights with (1) the current interpretive duty but (2) without the
power (perhaps expressly prohibited) to declare statutes incompatible with protected
rights. In practice, however, this might lead to even greater pressure on the courts to
distort parliamentary intent than with the current outlet of the declaratory power.
Finally, there is the issue of judicial appointments. One wonders how much
longer highest court judges in countries adopting the new model can continue to be
appointed in the traditional, comparatively non-political, common law manner -- by
the government alone based primarily on professional reputation, peer review, and
seniority -- before calls for the types of greater indirect political accountability and
legislative input, perhaps even the relevance of political affiliation, typical of
constitutional courts exercising de jure supremacy around the world become
unanswerable. Indeed, the UK has recently moved even further in the direction of
greater insulation from political factors (but also greater transparency) by instituting
a new, fully independent Judicial Appointments Committee to replace the opaque
method of selection by the Lord Chancellor. Predictably, given the Charter’s greater
age and grant of judicial power, Canada has already come under the most pressure on
162
CAROLINE EVANS AND SIMON EVANS, AUSTRALIAN BILLS OF RIGHTS (2008); Roger
Masterman, Interpretations, Declarations and Dialogue: Righrs Protection under the Human Rights
Act and Victorian Charter of Human Rights and Responsibilities, 2009 PUB. L. 112.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 40
this front and in 2006 held its first ever – albeit brief and entirely non-partisan --
public hearing for a nominee to the Supreme Court.163
In any event, increasing the
political – and perhaps especially legislative -- role in judicial appointments to the
highest courts exercising weak-form judicial review, as well as possibly also
adjusting tenure (perhaps renewable fixed terms in the highest courts), may also be a
means of recalibrating judicial and legislative power under the new model in favor of
the latter – if thought necessary, appropriate, and not counterproductive.
CONCLUSION
In reassessing the new model in light of practice, this article has advanced
three theses. First, the UK and New Zealand versions are working reasonably well
with respect to both protecting rights and balancing judicial and legislative power.
Second, from the perspective of the new model as a whole, the Canadian version
suffers from a serious practical problem due to the near dormancy of section 33 and,
however section 1 does or might operate, it is not a functional substitute. Third, one
principal dilemma has emerged from experience under the new model, to which I
suggest the practical solution of the legislative remedy.
In thus concluding that, overall, the new model has so far been moderately
successful and distinctive in operation, this article has engaged and challenged the
major expressed skeptical claims to the effect that practice is not living up to theory.
This theory, again, is that the new model promotes two constitutional goods or
foundational values of liberal democratic polities – adequate rights protection and an
appropriately balanced allocation of governmental powers, including limits on both
courts and legislatures. This article is not claiming, however, that the new model is
the best or universally appropriate model of constitutionalism.164
Rather, it is
claiming that the new model should be afforded its rightful place on the menu of
options as a third, intermediate alternative to judicial and legislative supremacy. No
doubt, as with both of the existing options, certain contextual factors –including the
presence or absence of cataclysmic change of regime and viewing one or other
model as more relevant for aspirational or historical reasons—will continue to play a
major role in the ultimate selection, but the choice may as a result be a more
informed one.
Finally, it should also be clear that the case for the new model is not simply
that it is intermediate. Obviously the fact of being in the middle, of being situated in
between two normatively supported options, does not itself provide a normative
argument for that position – except perhaps for unreconstructed Aristotelians. That
said, the normative claims that I have briefly restated and assessed in practice for
fulfillability in this article are not entirely independent of the new model’s
163
Although the process was not repeated when the next vacancy arose, in 2008.
164
Although premised on the traditional two choices only, I am generally persuaded by Wojciech
Sadurski’s argument that context is important on the issue of which type of legal regime best protects
rights. See Sadurski, supra note 13.
April 2010 Reassessing the New Commonwealth Model of Constitutionalism 41
intermediate status in important ways. For, at bottom, much of its appeal is that it
transcends two either/or options – judicial or legislative supremacy, and a legal or
political conception of rights -- previously thought to exhaust the institutional forms
of constitutionalism.