1

Constitutional conventions
b r i a n ga l l i g a n an d s c ot t b ren to n

This opening chapter addresses the foundational questions of what is a

convention, what gives a convention its legitimacy and how conventions
remain stable yet adaptable. Conventions are often defined negatively,
in terms of not being law, or in political terms with the implication that
they are inferior to law. We contend that conventions occupy a more
fundamental realm: government and the rules for forming government
precede law and make law-making possible. In other words, conventions
are more fundamental than laws; they govern the formation and basic
functioning of government overall and in its key parts, and governments
set up the law-making institutions that make, interpret and enforce laws.
Hence it is not appropriate to go backwards, as it were, and view
conventions from the more precise lens of law, which has been a strong
tendency since the nineteenth century.
Hence our starting point is consideration of the origins of systems
of government and law, and why they are supported and obeyed.
We propose a political approach to explain how conventions derive from
fundamental political principles that inform practice. After all, conven-
tions are basic political rules affecting the structure and powers of
government that are not enforceable in courts of law, so not amenable
to jurisprudential treatment that purports to view them as quasi-laws.
While for the most part conventions are not formally codified as laws are,
they are broadly accepted as binding by government actors and citizens.
They shape government practice albeit with flexibility that allows for
development and change.
As a crucial part of Westminster-derived constitutional systems,
conventions govern the institutions and operation of most aspects of
parliament and responsible government: the offices of prime minister
and cabinet, the rules for forming government, the dissolution of parlia-
ment, the appointment and removal of ministers, the role of the Queen
and vice-regal representatives, doctrines of collective and individual
ministerial responsibilities, the transaction of parliamentary business,
8

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 constitutional conventions 9
caretaker government, the independence of the judiciary, relationships
with and between the public service, the executive and the legislature and
the chain of accountability involving public servants. Conventions are
crucially important in countries with ‘unwritten’ constitutions like those
of the United Kingdom and New Zealand, and define the executive parts
of partly written constitutions such as those of Canada and Australia.
Conventions attracted more attention in the late nineteenth and
early twentieth centuries when parliamentary politics were more fluid
and vice-regal offices in Westminster dominions, as they were then
termed (Evatt 1936), were filled by UK appointees who played a more
active role. The rise and dominance of disciplined party politics for
much of the twentieth century usually produced clear electoral outcomes
with the vice-regal office, now filled by prominent domestic appointees,
becoming more formal. In recent years, however, all four countries have
experienced coalition and minority governments with multi-party cab-
inets, while Canada has a long history of minority government despite a
majoritarian electoral system.
Australia has been home to many non-traditional Westminster experi-
mentations, most notably the powerful Senate, and independents and
minor party parliamentarians have served in state cabinets. New Zealand,
once described as ‘more Westminster than Westminster’, has radically
changed its electoral system and fostered more consensus-style govern-
ment. There are continuing calls for democratic reform in the United
Kingdom and Canada, particularly in the upper house, while both coun-
tries also grapple with how to accommodate nationalist and separatist
movements. Democratic reform, at least for some (Marsh and Miller
2012), requires a multi-party system to better represent today’s more
complex postmodern society.
While the United States adopted a republican form of government,
with a full separation of powers, codified constitutional limitations
and democratic mandates for executive and legislative institutions and
actors, the United Kingdom and subsequently Canada, Australia and
New Zealand relied on conventions to perform similar functions. Con-
ventions were used to curtail executive power, formally the prerogative of
the monarch, and to regulate executive–legislative relations. In recent
decades there have been consistent calls for greater democratisation and
enhanced oversight and scrutiny of the executive in all four countries.
Implicit is a critique of responsible party government, with the major
political parties, traditionally the guardians of the ‘rules of the game’,
being challenged by new political actors.

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 10 brian galligan and scott brenton
The United Kingdom and New Zealand, with unitary systems of
government and without entrenched written constitutions, have been
leading the charge. New Zealand adopted a new system of voting,
MMP, that produces multi-party government, and has pioneered the
development of a comprehensive Cabinet Manual to explain how con-
ventions are adapted to its consequences, as Grant Duncan explains in
Chapter 12. The United Kingdom has been undergoing a ‘twenty-year
constitutional revolution’, according to Robert Hazell in Chapter 9, that
has entailed ‘the creation of new conventions; the modification of old
conventions; the codification of many conventions; and the demise of
conventions rendered obsolete.’ The intensity of discussion and develop-
ment of conventions has been more subdued in Canada and Australia,
except for flash points such as Canada’s prorogation crisis in 2008 and
Australia’s constitutional crisis of 1975. In these two countries, written
federal constitutions, combined with judicial review, structure and over-
see legislative and executive power to a far greater extent.
Despite different trajectories, the four countries under consideration
have experienced common changes. The most significant changes have
been driven by the challenges of forming governments and achieving
executive unity sufficient to ensure confidence while allowing for differ-
ences. Where governments have been formed through formal agree-
ments, there has also been a steady extension in the scope of policy
commitments and significant budget implications. The ‘logic of appro-
priateness’ has guided this process, and there is an expectation that
judgements, actions and decisions will have to be justified to others (see
March and Olsen 1989). This historical institutionalist account helps to
explain why many conventions have remained so stable. Political actors
have responded to various situations in what they considered was most
appropriate with regard to their position and responsibilities (Koeble
1995). The most appropriate response in a liberal democracy is generally
the one that ensures political stability, but also takes into account the will
of the people. For example, the major parties will look for a response that
preserves their dominant position while also acknowledging changed
political circumstances. Many scholars have described the cartel-like
behaviour that major political parties have engaged in (see Katz and Mair
1995), with the longer term objective of alternating in government.
While aspects of all of these have been considered by various authors,
the field as a whole is relatively neglected: the last authoritative book was
Geoffrey Marshall’s Constitutional Conventions (1984). Most commen-
tary on conventions is by constitutional lawyers who typically work from

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 constitutional conventions 11
a legal perspective, for example from Dicey, who first made the sharp
distinction between law and convention, to the present day (e.g. Sirota
2011). A legal perspective tends to restrict and often distorts the charac-
ter of conventions. Our approach is primarily a political one: drawing on
political theory, political history and accepted practice in determining
the nature of conventions and how they operate. We view constitutional
conventions as integral parts of the constitutional system that has
broad political and social, rather than legal, foundations (Galligan and
Versteeg 2013).
Although constitutional law and the courts that interpret and apply it
have a different institutional character, they are also part of that same
larger constitutional system. Constitutional law and conventions are not
the same, but differ by ‘a matter of degree’, as Barber (2009: 309) points
out: ‘Constitutional conventions and laws are two brands of social rule
which differ in the extent of their formalization.’ So rather than defining
conventions in negative terms as being ‘not law’, we see them as political
institutions that are core parts of the constitutional system and govern
ongoing political practice.
This first chapter sets out the basic theory and framework of conven-
tions, and outlines their character and the crucial role they play in
parliamentary and responsible government in established Westminster
systems. It will provide a framework for the overall understanding of
conventions. Major institutions, such as the constitutional monarchy,
cabinet government, upper houses or caretaker conventions, or develop-
ments, such as minority and multi-party governments or codifying
conventions in handbooks and cabinet manuals, are discussed in later
chapters. We begin with A.V. Dicey’s classic view that dominated
twentieth-century thinking about conventions and continues to influence
constitutional lawyers.

The legal view of conventions: Dicey’s legacy

A.V. Dicey, the Vinerian Professor of English Law at Oxford, published
the Law of the Constitution in 1885, revising subsequent editions over the
next thirty years to take account of major developments in UK and
Commonwealth governments. The eighth edition was published in
1914, with subsequent editions since his death. Although Dicey has been
roundly criticised by political scientists (e.g. McLean 2010), he has cast a
long shadow over constitutional thinking in Westminster systems.

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 12 brian galligan and scott brenton
Dicey focused on three leading principles: the sovereignty of parlia-
ment, the rule of law and the conventions of the constitution (Dicey
1973). Dicey’s treatment of all three of these leading principles was from
a legal perspective. The sovereignty of parliament was the foundational
principle recognised by courts, and the laws passed by parliament were
supreme. The parliament was from a legal point of view the absolute
sovereign of the British Empire, and every Act of Parliament was binding
on courts throughout its dominions.
These sweeping claims were a legal fiction – albeit a powerful one – in
Dicey’s time, and hardly captured the autonomy of the increasingly
independent nations that once were colonial dominions. Nevertheless,
the fiction of parliamentary sovereignty in the United Kingdom con-
tinued to have legal currency (Goldsworthy 2010), in part because of its
restraint in attempting to exercise such sovereignty and in part through
a curious transference of sovereignty thinking and discourse to the
parliaments of Canada, Australia and New Zealand. The UK parliament
had established the constitutional systems of those countries through
exercising its sovereign power, it was said, and the national parliaments
so established became parliamentary clones in exercising sovereignty
in their domains. The sovereignty of dominion parliaments was
restricted in leaving much of foreign policy making to the imperial UK
parliament, and for federal countries like Canada and Australia domestic
sovereignty was said to be divided between national and provincial or
state parliaments.
Not surprisingly in this world of legal parliamentary sovereignty and
the rule of law, constitutional conventions were only second best. Dicey’s
mature view is summed up in his answer to the question ‘Have there
been during the last thirty years notable changes in the conventions of
the constitution?’ This question is posed in his opening discussion
of the conventions of the constitution in the eighth edition of his book.
His answer distinguished two sorts of conventions:
namely, first, new rules or customs which still continue to be mere
constitutional understandings or conventions, and, secondly, understand-
ings of conventions which have since 1884 either been converted into laws
or are closely connected with changes of law. These may appropriately be
termed ‘enacted conventions’.
(Dicey [1885] 1973: lxxvi)

So here we have it: mere conventions or customs if they remain as such,

or otherwise conventions that have become law or are closely connected

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 constitutional conventions 13
to laws. This view colours Dicey’s answer to the subsequent question as
to why conventions are obeyed and how they are enforced. His answer is
not because of political consensus or the power of public opinion but
‘nothing else than the force of the law’. The sanction that constrains the
‘boldest political adventurer to obey the fundamental principles of the
constitution and the conventions in which these principles are expressed,
is in fact that the breach of those principles will almost immediately
bring the offender into conflict with the Courts and the law of the land’
(Dicey [1885] 1973: 445). This is clearly not the case as our subsequent
political account of conventions makes clear. Nor is there the sharp
distinction between conventions that are customs or ‘mere constitutional
understandings’ and ‘enacted conventions’ that become law. In Chapter 2
Nicholas Aroney provides a far more nuanced and complex account
of law and convention.
Dicey acknowledged the fact that ‘political sovereignty’ based on the
will of the electorate or nation underpinned both the supremacy of the
elected House of Commons and the rule that gives the appointment and
control of the government mainly to the House of Commons, but treated
these as ‘understandings’ or ‘customs’. They comprise the modern code of
‘constitutional morality’ that ‘secures, though in a roundabout way, what
is called abroad the “sovereignty of the people” ’ (Dicey [1885] 1973: 431).
Dicey influenced the constitutional thinking in Westminster countries
for generations, particularly among constitutional lawyers who domin-
ated the study of constitutions. Their focus was on constitutional law and
court decisions, viewing conventions as customs or political practices
that had inferior status. For the most part conventions were neglected
by political scientists as being part of a constitutional lawyers’ domain.
To properly base conventions, we need to take a more realistic political
approach that puts conventions at the core of the constitution with the
same basic foundation as constitutional law, both of which are inter-
related parts of the constitutional system. To use Dicey’s terms we need
to go beyond accepting the legal sovereignty of parliament and probe
the basis of its political sovereignty. This requires an account of
Westminster-style constitutional systems that have parliament and
responsible government as primary institutions.

The political view of conventions

The political approach to conventions draws upon the larger political
view of constitutional systems that pays primary attention to the nature

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 14 brian galligan and scott brenton
of constitutions as systems of government that perform the complex
process of governing and are accepted by the people, political elites and
citizens, in which they operate. Accordingly, the UK constitutional
system has at its core the primacy of Parliament, and in modern times,
of the House of Commons as the elected house of parliament. Dicey’s
foundational law principle whereby the courts acknowledge parliamen-
tary sovereignty is the legal recognition of this political state of affairs.
There is no prior legal principle or rule that established such sovereignty;
nor does the parliament have absolute sovereignty, but rather a politically
limited authority and jurisdiction. We might restate these two key points
in positive terms: UK parliamentary sovereignty is based on convention,
and it is limited by political factors. The evidence for these two propos-
itions is realistic and historical.
The primacy or legislative authority of parliament is based on its
general acceptance within the polity of the United Kingdom. This fact
is well recognised by the prominent legal scholar H.L.A. Hart, who states
that general acceptance is ‘a complex phenomena, in a sense divided
between official and ordinary citizens, who contribute to it and so to the
existence of a legal system in different ways’. Hart explained further:
The officials of the system may be said to acknowledge explicitly such
fundamental rules conferring legislative authority: the legislators do this
when they make laws in accordance with the rules which empower them
to do so: the courts when they identify, as laws to be applied by them,
the laws made by those thus qualified, and the experts when they guide
the ordinary citizens by reference to the laws so made. The ordinary
citizen manifests his acceptance largely by acquiescence in the results of
these official operations. He keeps the law. . .
(Hart [1961] 1972: 59/60)

Evidence for the existence of a legal system, Hart argues, depends on

the congruence or acceptance of ‘two different sectors of social life’, the
official and the citizens: ‘Crudely put, the facts are that the rules recog-
nized as valid at the official level area are generally obeyed’ (Hart [1961]
1972: 114). This is the ultimate ‘rule of recognition’ on which legal
systems are based.
While Hart is discussing the foundations of a legal system, his obser-
vations that its existence depends on the recognition and acceptance of
two groups, officials and citizens, applies more generally to constitutional
systems. So also does his example of ‘the embryology of legal systems’ in
the Commonwealth countries. As Canada, Australia and New Zealand
became progressively autonomous nations with their legal systems, the

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 constitutional conventions 15
legislative authority by the Westminster Parliament was reduced and
finally terminated. This was achieved through a series of Commonwealth
conferences with decisions formalised in various declarations and legis-
lation at Westminster and in the national parliaments, and recognised by
the courts. Legal recognition followed and formally registered political
achievement, even if there were unresolved legal conundrums such as a
supposed sovereign Westminster Parliament having its sovereignty
reduced.
Scottish devolution, and possible secession, is a case where Westmin-
ster’s authority has been restricted by the political necessity to give
autonomy to Scotland and the Scottish Parliament to manage Scottish
affairs. While Scottish people recently voted not to separate from the
United Kingdom, the decision was theirs to make. It is no longer the case
that an oligarchic few could change Scotland’s status through an act of
union, and the people acquiesce. Nor does a legalist perspective, albeit the
‘soft law’ one of Aileen McHarg, provide a plausible narrative. In her
account that relies on traditional notions of the legal sovereignty of
Westminster,

the Sewell convention gives some recognition to the Scottish Parliament’s

claim to political sovereignty whilst maintaining the legal supremacy of
the Westminster Parliament, just as decades earlier the conventions
declared at successive Imperial Conferences acknowledged the growing
political equality of the United Kingdom and its Dominions whilst
retaining Westminster’s formal status as an Imperial Parliament’.
(McHarg 2008: 868)

Exaggerated notions of legal sovereignty and supremacy are retained with

political reality allowed in as a soft law qualifier. It is preferable to
acknowledge the primacy of changing political reality and constitutional
systems, and seek appropriate legal terminology for their formal
recognition.
Constitutional systems together with their component conventions
and legal systems are political arrangements. In Russell Hardin’s mega
conventional view that he draws from the eighteenth-century political
philosopher David Hume: ‘Government derives its power (not its right)
to rule by some specific form of coordination that is a convention and
the populace acquiesces in that rule by its own convention’ (Hardin 2013:
59). In other words, the whole constitutional system, including its legal
and conventional parts, is a particular sort of grand coordinating system
that is conventional in form. That it is a system of government requires

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 16 brian galligan and scott brenton
the second convention of recognition and acceptance or acquiescence by
those within its jurisdiction. Courts and law are part of this constitutional
system, albeit with distinctive organisational form and rules of operation.
Conventions work in ordering the parliamentary legislature constitut-
ing the executive government in Westminster parliamentary systems
because they are broadly supported, accepted or acquiesced to by both
political elites and the people, or at least those who are more engaged
and active citizens. That conventions work in a routine but flexible way
in shaping legislative and executive government is hardly contestable
because the four parliamentary democracies under consideration, the
United Kingdom, Canada, Australia and New Zealand, are among
the most stable, enduring and well-governed polities. In these countries
it is clearly apparent that conventions operate and are broadly supported
by political elites and the citizenry. There is some debate in the literature
about the character of popular support: whether it really is strong
support or weaker acquiescence on the part of the people. For Russell
Hardin, ‘acquiescence is the compelling fact’ for grounding ‘the dual-
convention theory of government’ (2013: 60; emphasis in original).
Hardin’s critical realism that draws on Hume rejects alternative con-
tract theories of government deriving from Thomas Hobbes and John
Locke that were prominent in Hume’s time, and still have currency today
(Ginsberg 2013: 182). Locke had written his influential Second Treatise of
Government to justify the Glorious Revolution of 1688 when parliamen-
tary oligarchs terminated the absolutist Stewart line in favour of a limited
monarchy. According to Locke’s account of the origins of government,
people who were naturally free and equal came together and agreed to
enter civil or political society, exchanging their natural rights for civil or
political rights. They agreed to the broad framework of government, and
were entitled to change it, or key parts of it, if it became tyrannical
and unsuitable for their interests. The formation of government was a
solemn agreement or sort of original contract in the first instance and
required at least ‘tacit’ consent for its continuation. Various forms of
contract theory informed much of eighteenth-century political thinking
and were influential in the constitutional founding of the United States.
As Locke made clear, his was a theoretical rather than a historical
account of the origins and basis of legitimate government. This was not
always acknowledged by subsequent contract theorists reflecting on the
constitutional convocations that drafted the American constitution
and the rhetoric of articulate founders like Jefferson in drafting the
Declaration of Independence that justified severing British rule, and

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 constitutional conventions 17
Madison in explaining and defending the new American constitution in
the influential Federalist Papers.
The critical realist view has an alternative account of constitutional
origins, and a different intellectual pedigree. Hume’s account of the
constitutional break that brought William and Mary to the throne in
1688 is one of elite action and popular acquiescence: ‘it was only the
majority of seven hundred, who determined that change for ten mil-
lions...[T]he bulk of those ten millions acquiesced willingly in the deter-
mination’ (Hume [1772/1777] 1994: 191). Hume did allow that the
consent of the people would be a just foundation of government, but
that it was rare and ‘never almost in its full extent’. Nevertheless, as the
case of switching monarchic succession to William and Mary showed,
constitutional change can be effected and become widely accepted
(Hume [1772/1777] 1994: 192).
Such pragmatic empiricism that acknowledges the fact of, and support
for, government and basic changes to government structures does need
some further explanation. Why do political elites and the people
governed continue to support or acquiesce in the system of government
that orders and limits their behaviour? Hume’s answer to this broad
question of allegiance and obedience to the law was ‘common interest
and utility’. ‘[I]f government were totally useless, it never could have
place’ Hume argued, ‘and that the sole foundation of the duty of alle-
giance is the advantage, which it procures to society, by preserving peace
and order among mankind’ (Hume [1777] 1972: 205, 211, emphasis in
original). If this seems a weak conclusion, it is in line with the earlier
tradition of Hobbes and Locke that people entered political society under
government in order to achieve individual security, peace and order.
At the end of the day, according to Hume, matters will likely be decided
by ‘an appeal to general opinion’: ‘there is really no other standard, by
which any controversy can ever be decided’ (Hume [1772/1777] 1994:
200). Hume might have added force of arms that often achieve or
facilitate, at least in part, the founding of political systems or major
changes within them.
Explaining the existence and endurance of systems of government has
been a major concern of political science since Renaissance times with
Machiavelli, the ultra-realist, finding the explanation in a combination
of political circumstances and the character of, and dynamic interrela-
tionship between, the rulers and people. To succeed, Machiavelli’s
Prince needed to tailor his behaviour to suit the times and the character
of the people – ruling for the common good if the people were virtuous,

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 18 brian galligan and scott brenton
or as a ruthless tyrant if they were corrupt. Similarly for republics,
that Machiavelli preferred, provided the people had the qualities that a
republic required, rulers needed to curb their ambitions for the public
good, as was the case in the ancient Roman republic before Caesar and
the emperors. This realist strand was also prominent in the American
constitutional founding that assumed the self-interested behaviour of
politicians but sought to channel and curb this through clever insti-
tutional architecture incorporating the division of powers and checks
and balances within and between the branches of government.
According to modern realists like Ran Hirschl, stakeholders abide by
constitutional constraints out of self-interest. Although Hirschl is dis-
cussing the foundation of constitutions, his observations might apply
equally to their ongoing support. Political actors who voluntarily estab-
lish institutions that appear to limit their institutional flexibility, Hirschl
says, ‘must assume that the clipping of their wings’ will be compensated
for by the limits imposed on rivals and the reduced risk of other non-
favourable political developments in the future (Hirschl 2013: 167).
Limiting political actors and channelling their actions are functions of
conventions to which we now turn.

Conventions that bind

Whether some practice is a convention or not is fundamental for under-
standing what conventions are and a threshold issue for projects pur-
porting to codify them. We might begin with two basic aspects:
conventions are principles and practices of government that are both
recognised and binding. The two aspects are necessary: recognition of
conventions among key political actors and commentators is an obvious
prerequisite; their acceptance as binding is also required if conventions
are to govern political activity. This dual requirement has been generally
recognised by leading scholars in the past. Sir Kenneth Wheare, the
Australian who was prominent at Oxford, defined conventions in his
Modern Constitutions: ‘By convention is meant a binding rule, a rule of
behaviour accepted as obligatory by those concerned in the working
of the Constitution’ (Wheare 1951: 27).
Wheare also reconsiders earlier classifications of constitutions in terms
of rigidity and flexibility (see Bryce 1884). While the ‘procedures’ for
changing conventions are not rigid like many written constitutions, in
practice they are not easy to alter due to the very absence of a tangible
and accepted means of immediate change.

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 constitutional conventions 19
Others have given a more extensive set of requirements that have to be
met in recognising conventions, with their binding requisite being an
integral part. Sir Ivor Jennings proposed a tripartite test that turned on
three questions, to which he gave partial answers:

first, what are the precedents; secondly, did the actors in the precedents
believe that they were bound by a rule; and thirdly, is there a reason for
the rule? A single precedent with a good reason may be enough to
establish the rule. A whole string of precedents without such a reason
will be of no avail, unless it is perfectly certain that the persons concerned
regarded them as bound by it.
(Jennings 1959: 136)

Jennings’ test is significant because it was adopted by the Canadian

Supreme Court in deciding the Patriation Reference case (1981: SCR
753). Paradoxically for a court of law, but in an advisory opinion, the
Supreme Court of Canada advised that there was a convention requiring
the federal government to have substantial provincial support before it
could properly advise the UK Parliament to make amendments to the
Canadian constitution affecting provincial powers. This occurred just
prior to the patriation of the constitution when the Trudeau government
sought to make amendments without provincial support. Regardless of
this weighty endorsement, Jennings’ test is hardly definitive and its
application depends on judgement and reasonable consensus (Heard
2014: 172ff.). All three parts of Jennings’ test are contestable: sorting
out legitimate precedents; agreeing on what is a good, or a good-enough
reason, to establish a rule; and being satisfied that actors believed they
were bound by a rule. In this case the Canadian Supreme Court by a
majority decision supported the federalist side in a political dispute about
a convention, and that was enough to carry the day.
Geoffrey Marshall, in his classic account of Constitutional Conventions
(1984), discusses both the recognition and binding character of conven-
tions at some length, but without any crisp resolution. Marshall argues
that Wheare’s ‘emphasis on obligatory behaviour . . . may obscure the
point that the conventions, as a body of constitutional morality, deal not
just with obligations or duties but confer rights, powers, and duties’
(1984: 7). In summarising his views, Marshall concludes that ‘No general
reason needs to be advanced to account for compliance with duty-
imposing conventions beyond the fact that when they are obeyed . . .
they are believed to formulate valid rules of obligation’ (Marshall 1984:
210). This does not take us very far in determining what are or are not

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 20 brian galligan and scott brenton
conventions. Saying that a convention requires belief on the part of
the observer that they are valid rules of obligation does not preclude
mistaken belief. Nor is it clear that we should preclude those where
observance is due to political calculation or rational choice rather
than belief.
What kind of obligation does a duty-imposing convention impose?
Marshall asks this in discussing Jennings’ test that actors must believe
they are bound by a rule. Marshall’s best answer is ‘the critical morality
of the Constitution’ that he prefers to the ‘positive morality of the
Constitution’ because it enables one to judge actual behaviour against
what ‘the political actors ought to feel obliged by’ (Marshall 1984: 11–12).
Although it follows Dicey and Jennings, this all seems to make the
discernment of conventions too much a psychological guessing game.
Do we need to know what political actors believe or ought to believe
about rules, rather than observing patterns of behaviour and political
consensus?
On the recognition of conventions, Marshall is clear that ‘If a conven-
tion is to be a guide to conduct it must be known what course of action it
prescribes’ (Marshall 1984: 211). But as he immediately acknowledges,
conventions are plagued by vagueness, and he instances the Canadian
convention upheld by the Canadian Supreme Court in the Patriation
Reference case. The Court’s decision added weight to the provincial case,
and no doubt tipped the political balance in their favour, although
dissenting judges took opposing views. In any case, Marshall points out
that conventions have a spectrum from clarity to vagueness that other
rules of behaviour share: ‘namely that what they require in some
clear cases is known but what they require in more marginal or arguable
cases cannot be stated in advance’ (Marshall 1984: 211). As the following
chapters show, there are multiple ways and instruments in which con-
ventions can be specified, even if some are contested.

Codification and constitutional reform

Contemporary discussions and debates about conventions often focus on
whether to codify and to what extent, which leads to broader issues of
constitutional reform. Codes are essentially written forms, and can
include codes of conduct, guides, manuals and even legislation. In this
volume we make a distinction between this latter form and the others.
Codification in written forms other than through legislation and the
constitution does not change the status of conventions. While it makes

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 constitutional conventions 21
conventions clearer, such formulations are provisional, contentious and
subject to adjustment and change. Yet these are the main forms of
codification, with only very few examples of political actors seeking
to legally or constitutionally codify conventions, unless they are part of
other constitutional reforms.
Rather, practitioners and scholars have generally been more interested
in forms of codification that can be easily updated, recognising that one of
the reasons for creating a code is to help clarify new and at times uncertain
situations. Codes articulate how conventions are viewed at a particular
time but need to evolve as conventions change. Many contemporary codes
are endorsed by each new government, and that provides an opportunity
to revise the code and reflect on its previous application (e.g. New Zealand
Cabinet Office 2008). However, this process vests power in the major
political parties, with other political actors only able to provide input into
the process if they are in an influential position during the government
formation phase. These factors are all important for understanding the
tension between stability and adaptability of conventions.
Calls for codification through legislative or constitutional means
generally arise when more significant constitutional reforms are being
proposed. These more significant constitutional reforms tend to be ones
that democratise, but, at the same time, weaken the political positions of
the older major parties. For example, devolution in the United Kingdom
and the introduction of the proportionally representative electoral system
in New Zealand have brought conventions to the fore, along with a
plethora of new political actors that see new political systems as an
opportunity to adapt practices that have unduly empowered the incum-
bents. In these cases, all actors have an interest in clarifying what the
practice should be, and legislative or constitutional change is seen to
be the best way of guaranteeing ongoing compliance. Codification in
New Zealand and the United Kingdom has been more common, where
change has been greater and constitutional entrenchment less than in
Canada and Australia. It is also no coincidence that these countries lack
written constitutions.
While in Australia and Canada it is more difficult to pursue substantial
legal or constitutional codification for many reasons, there is also a
particular deference to the written constitution and often debates will
centre on deciphering the original intentions of the founders. Major
crises such as that of Australia in 1975 and Canada over prorogation in
2008 leave little hope of resolution because of the clash of conventions
and ongoing dispute over their application. Yet in these countries,

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 22 brian galligan and scott brenton
proposals for constitutional change still show the inextricable link
between democratisation and clarifying conventions through codifica-
tion. In Canada, such issues include the appointed upper house, accom-
modating diverse interests in the federation and multi-party power
sharing. In Australia, moves towards a republic have raised questions
about the mandate of the head of state (whether or not to directly elect)
and the reserve powers of the current governor-general. In the following
chapters we consider common issues across the four countries, while also
focusing on the unique features of each country.

References
Barber, N.W. 2009. ‘Laws and Constitutional Conventions’, Law Quarterly Review
125: 294–309.
Bryce, James 1884. Constitutions. New York: Oxford University Press American
Branch.
Dicey, A.V. [1885] 1973. Introduction to the Study of the Law of the Constitution,
10th edn. London: Macmillan.
Evatt, H.V. 1936. The King and His Dominion Governors. Oxford University Press.
Galligan, Denis and Versteeg, Mila (eds.) 2013. Social and Political Foundations of
Constitutions. Cambridge University Press.
Ginsberg, Tom 2013. ‘Constitutions as Contract, Constitutions as Charters’, in
Galligan, Denis and Versteeg, Mila (eds.), Social and Political Foundations of
Constitutions. Cambridge University Press, 182–204.
Goldsworthy, Jeffrey 2010. Parliamentary Sovereignty: Contemporary Debates.
Cambridge University Press.
Hardin, Russell 2013. ‘Why a Constitution?’, in Galligan, Denis and Versteeg, Mila
(eds.), Social and Political Foundations of Constitutions. Cambridge Univer-
sity Press, 51–72.
Hart, H.L.A. [1961] 1972. The Concept of the Law. Oxford: Clarendon Press.
Heard, Andrew 2014. Canadian Constitutional Conventions: The Marriage of
Law & Politics, 2nd edn. Oxford University Press.
Hirschl, Ran 2013. ‘The Strategic Foundations of Constitutions’, in Galligan, Denis
and Versteeg, Mila (eds.), Social and Political Foundations of Constitutions.
Cambridge University Press, 157–81.
Hume, David. [1772/1777] 1994. ‘Of the Original Contract’, Essay Twenty-Two,
in Haakonssen, Knud (ed.), Political Essays. Cambridge University Press,
186–201.
Hume, David. [1777] 1972. ‘Of Political Society’, Section IV, in Selby-Bigge, L.A.
(ed.), Enquiries Concerning the Human Understanding and Concerning the
Principles of Morals.. Oxford: Clarendon Press, 205–11.

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002
 constitutional conventions 23

Jennings, Sir Ivor 1959. The Law and the Constitution, 5th edn. University of
London Press.
Katz, Richard S. and Mair, Peter. 1995. ‘Changing Models of Party Organization
and Party Democracy’. Party Politics 1(1): 5–28.
Koeble, Thomas A. 1995. ‘The New Institutionalism in Political Science and
Sociology’. Comparative Politics 27(2): 231–43.
March, James G. and Olsen, Johan P. 1989. Rediscovering Institutions: The Organ-
izational Basis of Politics. New York: Free Press.
Marsh, Ian and Miller, Raymond 2012. Democratic Decline and Democratic
Renewal. Cambridge University Press.
Marshall, Geoffrey 1984. Constitutional Conventions. Oxford: Clarendon Press.
McHarg, Aileen 2008. ‘Reforming the United Kingdom Constitution: Law,
Convention, Soft Law’, Modern Law Review 71: 853–77.
McLean, Iain 2010. What’s Wrong with the British Constitution. Oxford University
Press.
New Zealand Cabinet Office 2008. Cabinet Manual. Wellington, New Zealand:
Cabinet Office, Department of the Prime Minister and Cabinet.
Patriation Reference 1981. Reference re: Resolution to Amend the Constitution.
1 SCR 753.
Sirota, Leonid 2011.‘Towards a Jurisprudence of Constitutional Conventions’,
Oxford University Commonwealth Law Journal 11: 29–51.
Wheare, K.C. 1951. Modern Constitutions. Oxford University Press.

Downloaded from https://www.cambridge.org/core. University of Melbourne Library, on 09 Nov 2019 at 18:57:20, subject to the Cambridge
Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781316178560.002