Home | Databases | WorldLII | Search | Feedback University of Melbourne Law School Research Series

Stephenson, S --- "When Constitutional Conventions Fail" [2015] UMelbLRS 1

Last Updated: 22 May 2019

When Constitutional Conventions Fail

Scott Stephenson

First published in (2015) 38(2) Dublin University Law Journal

I. Introduction

Quiet concern is an underlying theme in much of the scholarship on constitu- tional conventions. While conventions may be a necessary feature of constitu- tional government, commentaries are often quick to point to the difficulties that arise due to the uncertainties that surround their content and enforce- ability. In this symposium, Adrian Vermeule’s paper contains the most explicit acknowledgement of these problems, claiming that the best the United States can do when it comes to the judicial recognition of constitutional conventions is to follow the ‘unhappy compromise’¹ found in the Commonwealth, given that it is ‘the least bad of the alternatives’.² Andrew Heard observes that, even though constitutional conventions are needed to ‘transform the positive law of the Constitution into a working system of government’,³ there are ‘prob- lems in either drawing from or ignoring them in judicial decisions’.⁴ His paper is directed to finding a workable means of distinguishing which conventions should and which conventions should not be judicially enforceable given that some are afflicted by ‘vagueness, controversy, and inconsistent observance’ while others are not.⁵ Beyond the symposium, a vast body of work exists attempting to come to terms with, and ultimately clarify, these uncertainties by creating tests to determine the existence and content of conventions as well as the extent to which courts can and should rely upon them.⁶

1. Adrian Vermeule, ‘Conventions in Court’ (2015) 38 DULJ 283, 285.
2. ibid 300.
3. Andrew Heard, ‘Constitutional Conventions and Written Constitutions: the Rule of Law Implications in Canada’ (2015) 38 DULJ 331.
4. Ibid 356.
5. Ibid 332.
6. The most well-known test for identifying conventions was set out by Ivor Jennings: see The Law and The Constitution (5th edn, University of London Press 1967) 136. For one criticism of Jennings’approach, see Andrew Heard,‘Constitutional Conventions: The Heart of the Living Constitution’ (2012) Journal of Parliamentary and Political Law 319, 332–37. See also Geoffrey Marshall, Constitutional Conventions: The Rules and Forms of Political Accountability (Oxford University Press 1984); Joseph Jaconelli, ‘The Nature of Constitutional Convention’ (1999) 19 Legal Studies 24; Joseph Jaconelli, ‘Do Constitutional Conventions Bind?’ (2005) 64 Cambridge Law Journal 149; Andrew Heard, Canadian Constitutional Conventions: The Marriage of Law & Politics (2nd edn, Oxford University Press 2014).

In this paper, I argue that constitutional conventions can also be a source of cautious optimism precisely on account of the uncertainties that surround their content and enforceability. Further, I contend that this dimension emerges when we might least expect, namely when constitutional conventions fail as a result of constitutional actors departing or threatening to depart from them. A failure can create a discursive moment, presenting constitutional actors, including the public, with a rare opportunity to discuss a neglected constitu- tional issue or to reframe a constitutional issue where debate has reached an impasse and become stagnant. During these moments, the doubts surround- ing their content and enforceability can be a positive influence because they encourage participants to engage directly with the convention and the justifi- cations for its existence. It is more difficult to avoid engaging with the merits of the issue by relying solely on past precedent if that precedent is strongly dis- puted or by shifting responsibility for resolving the issue to a court if questions surround that court’s authority to hear the case. These moments are a cause for cautious optimism because they only generate opportunities. Opportunities can be seized or squandered and, if seized, can result in productive or destruc- tive change. They place a previously ignored issue on the table or cast a new light on a pre-existing issue, but the quality of the outcome will ultimately depend on the choices the participants make.
This argument is developed in three stages. In Part II, I contend that the sta- bility some constitutional orders enjoy can have a negative side effect, namely stymying and distorting discussions about the system of government – its merits and possible reforms to it. Constitutional discourse founders as people (except perhaps for dedicated constitutional scholars) settle for the present arrangements because reform is considered to be too difficult and not suf- ficiently urgent. In Part III, I argue that the failure of a constitutional conven- tion can be a sufficiently serious event to shake constitutional actors from this view, forcing participants to confront the problem, to consider the options for change, and to expend the political capital necessary to bring questions about possible reforms to a head. From the disruption and damage caused by the failure of a constitutional convention comes an opportunity for constitutional reconsideration and renewal.
In Parts IV and V, I provide two examples where the failure of a constitutional convention produced a deliberative moment. Beginning with Australia, the Governor-General’s decision in 1975 to break with convention and not follow the advice of the Prime Minister provided an opportunity to debate two issues that warranted attention, but that had never been addressed in a detailed, sustained manner. The first is the Senate’s role in a system of responsible government, which arises because the Australian Constitution fuses elements from both the American and British models of government. Although the House of Representatives is the chamber where governments are formed and defeated (similar to the UK), the Senate has considerable power over legislation proposed by the government (sim- ilar to the US), giving rise to questions about when, if ever, it is appropriate for the Senate to take actions that might bring down the government. The second issue




concerns the appointment, identity and powers of the country’s head of state — the Governor-General as the Queen’s representative. This subject had come to merit reconsideration after Australia acquired independence from the United Kingdom over the course of the 20th century. While the opportunity presented by the events of 1975 to consider the first issue was largely squandered, it was affir- matively seized in respect of the second issue.
Moving to Canada, the Prime Minister’s proposal in 1980 to break with con- vention and proceed with amendments to the Constitution irrespective of the consent of the provinces presented an opportunity to reframe the longstand- ing debate on patriation and, more generally, the country’s federal system. For many decades, discussions on patriation (the creation of a procedure for amending the Constitution not involving the UK) had resulted in similar pro- posals on the assumption that any proposal required the consent of all the provinces.⁷ Once the Prime Minister proposed to break with past practice and proceed without any provincial consent, it was possible to reshape the terms of debate with the provinces having to negotiate under the threat that, if they were too obstructionist, the federal government would act unilaterally. Patriation was achieved in 1982. The opportunity presented by the failure of the constitutional convention was therefore seized, but it came at a heavy price. Quebec strongly opposed the manner in which the amendment procedure was introduced, destabilising the constitutional system with its protests and moves towards independence, forcing the country to seek on two occasions (both ulti- mately unsuccessful) to propose further amendments to appease Quebec.

II. The Downsides of Constitutional Stability

Some countries have constitutional orders that are exceedingly stable.⁸ They have systems of government that have existed for extended periods of time – one, two or perhaps more centuries – without fundamental revision. On the basic questions of government, they return the same answers that they did at their point of creation, remaining a bicameral (or unicameral), presidential (or parliamentary), federal (or unitary) monarchy (or republic) with (or without) a bill of rights and with broadly the same judicial structure and electoral system.⁹

7. The Supreme Court of Canada subsequently held by a narrow majority that the con- vention only required the consent of ‘a substantial majority’ of the provinces, but up until that decision negotiations had proceeded on the assumption that unanimous provincial consent was required: see text accompanying n 38.
8. They are, however, the exception rather than the rule: ‘most constitutions die young, and only a handful last longer than fifty years’: Zachary Elkins, Tom Ginsburg and James Melton, The Endurance of National Constitutions (Cambridge University Press 2009) 1.
9. In some instances, the answer that continues to be returned is silence. There are, for example, few references to the electoral system in the Australian and Canadian Constitutions.

They are places where ‘the basic constitutional arrangements are’, as Vermeule observes, ‘no longer up for grabs’.¹⁰
Australia and Canada are two countries that exhibit a high degree of consti- tutional stability. Since Australia became a federation in 1901, there has been no major change to its system of government.¹¹ The Constitution has been amended a mere eight times, with none fundamentally recasting its basic structure.¹² Canada has seen one major change to its system of government in 1982, which is the subject of discussion in Part V of the paper, when a bill of rights was intro- duced and the amendment procedure was significantly overhauled. Between federation in 1867 and patriation in 1982, there were 22 amendments made to Canada’s main, but not exclusive, constitutional document, the most significant of which expanded the pre-existing federal system through the creation or inclu- sion of new provinces.¹³ While important changes related to government have occurred in both countries during these periods such as independence from the UK and the development of the administrative state, the constitutional order has accommodated them either through small, incremental adjustments¹⁴ or through statutes designed to supplement the Constitution.¹⁵ Save for the events of 1982 in Canada, the constitutional narratives of both countries are primarily ones of continuity and constancy rather than discontinuity and disruption.
Their stability is extraordinary even by the standards of their close constitu- tional relatives. New Zealand and the United Kingdom are two other countries

10. Adrian Vermeule, Mechanisms of Democracy: Institutional Design Writ Small (Oxford University Press 2007) 2.
11. There have, however, been important structural changes at the subnational level, for example, the abolition of the upper legislative chamber in Queensland (Constitution Amendment Act 1921 (Qld)) and the introduction of bills of rights in the ACT (Human Rights Act 2004 (ACT)) and Victoria (Charter of Human Rights and Responsibilities Act 2006 (Vic)).
12. The successful amendments have concerned minor issues (1906, 1910, 1928), expan- sion of the federal government’s power to legislate with respect to social services (1946) and indigenous persons (1967), removal of discriminatory language against indigenous persons (1967), changes to the procedure for filling casual vacancies in the Senate (1977), the right of residents of the Territories to vote in referendums (1977), and the imposition of a retirement age for federal judges (1977).
13. A list of the amendments can be found in Guy Favreau, The Amendment of the Constitution of Canada (Canada Ministry of Justice 1965) 5–7. Some of the other doc- uments that formed part of the Canadian Constitution prior to 1982, and continue to do so today, are explicitly incorporated by way of the 1982 amendments: see Constitution Act 1982, s 52(2)(b).
14. See, eg, the amendments that expanded the federal government’s powers to legis- late on matters related to the provision of social security in Australia (1946) and Canada (1940, 1951 and 1964).
15. See Scott Stephenson,‘The Rise and Recognition of“Soft” Constitutional Law’, paper presented at The ‘Soft’ and The ‘Fuzzy’ in International and Public Law, Hebrew University of Jerusalem, Jerusalem, 17 February 2015 (on file with author).

with highly stable constitutional orders. Indeed, Vermeule identifies the latter as a prime example of a system where ‘the basic constitutional arrangements are no longer up for grabs’.¹⁶ Yet New Zealand eliminated its upper legislative chamber in 1950, introduced a bill of rights in 1990 and switched to a system of proportional representation in 1994, while the United Kingdom substantially reduced the power of the House of Lords in 1911 and again in 1949, enacted a bill of rights in 1998 and devolved power to Scotland and Wales in 1999. A key explanation for the difference between these two sets of countries is that New Zealand and the United Kingdom have uncodified constitutions, meaning that it only takes an ordinary statute to undertake major constitutional revision. By contrast, in Australia an amendment requires approval by both Houses of federal Parliament and by a majority of voters nationally as well as a majority of voters in a majority of states at a referendum. As will be discussed below, prior to 1982 Canada’s constitutional amendment procedure was the subject of some debate, but ordinarily involved the unanimous agreement of the federal and provincial governments for any change directly affecting the provinces.
Stability is, of course, a considerable virtue, as it both fosters an environ- ment for government to function well and is an indicator of a well-functioning government.¹⁷ It is, for instance, conducive to creating a predictable working environment for government officials and a predictable set of mechanisms for the public to access and participate in the operations of government and for them to hold it to account. Predictability in turn allows government officials and members of the public to focus on actions that are not directly related to the constitution — to concentrate on the quotidian work of government such as the provision of public services — and to take actions that are designed to yield medium- and long-term benefits.
Stability is, however, not without its vices. It can induce complacency where a country fails to critically reflect on its system of government and consider options for improvement. It will typically be easier for government actors to make do with what they have, begrudgingly accepting any problems or developing patchwork solutions to them, than to spend the considerable political capital it would take to initiate, persist with and ultimately achieve major reform. Attempts to under- take structural reform might be considered exercises in futility given that differ- ent groups will prefer different solutions, the task will distract from the quotidian matters of government and there may be little reward from the general public. The result is a growing catalogue of constitutional deficiencies — gaps, inconsis- tencies, anachronistic provisions — as the preconditions and rationales for the pre-existing set of arrangements dissipate with the changing economic, political and social environment. Discussions on an issue are likely to be especially difficult

16. Vermeule, Mechanisms of Democracy (n 10) 2.
17. For an extended discussion of both the virtues and vices of constitutional stability, see Elkins, Ginsburg and Melton (n 8) ch 2.

to initiate and maintain if that issue is anchored, even partially, in a text that is entrenched and thus burdensome to amend.

III. Constitutional Convention Failure as an Opportunity

In a highly stable constitutional order,¹⁸ advocates for reform must wait for a suit- able opportunity to put an issue on the political agenda or, if an issue is already recognised in some form, to advance debate by reshaping the terms of inquiry. Different types of events present different types of opportunities. An economic or political event might prompt investigation into whether the pre-existing sys- tem of government is suited to responding to it. The Great Depression, for exam- ple, was instrumental in prompting Canada to establish the Royal Commission on Dominion-Provincial Relations in 1937 to investigate whether its federal sys- tem was configured in a way that would allow it to respond to the changing economic environment. A political movement may successfully campaign the government to investigate a constitutional issue. In 2012, the Australian govern- ment established the Joint Select Committee on Constitutional Recognition of Aboriginal and Torres Strait Islander Peoples in response to growing pressure for constitutional recognition of indigenous peoples. A significant anniversary may also motivate the government to initiate an inquiry into the constitutional order, as occurred in Australia with the Constitutional Commission that issued its final report in 1988, which was the Bicentenary of British settlement.
The failure of a constitutional convention is another such event. Indeed, it is a particularly powerful prompt for constitutional reflection because it is both internal and urgent. A change to the environment in which the constitution operates such as an economic crisis may have multiple causes and solutions, many of which may be unrelated to the constitutional order and cannot be resolved by making changes to it. By contrast, failure of a constitutional con- vention is a failure of the constitutional order. It shines a light directly on a gap between the constitution’s formal arrangements and institutional practice, challenging actors to consider whether that gap can and should be remedied through changes to either the formal arrangements or the institutional prac- tice. Unlike a political movement or a significant anniversary, failure of a con- stitutional convention has an urgency that operates as a powerful motivating factor for government actors to discuss the issue with an eye to change. The crisis or controversy that the failure generates can help imbue participants with the mind-set that reforms, and thus the debates preceding reform, are essential rather than merely desirable. As constitutional change is often a long and slow process that can take years if not decades, that crisis or controversy can help keep the deliberative process energised and infused with a sense of purpose.

18. For the purposes of this paper, I am putting to one side events that disrupt consti- tutional stability to such an extent that they are prone to prompt complete constitu- tional overhaul such as revolution or foreign occupation.

As with all events, the opportunity these failures generate can be squan- dered or seized. In the immediate aftermath, debate may devolve into volleys of personal recriminations, obscuring the constitutional issues that precipitated or exacerbated the failure. Some actors, especially those that benefitted from the failure, may attempt to ignore the event or devalue its broader significance, urging the public to focus on the future rather than linger in the past. And when actors do seize the deliberative moment, we should not presume that formal constitutional amendment will occur. Campaigns for reform may lose momen- tum or participants may conclude that the failure was an aberration that therefore does not warrant any specific remedial step. For the purposes of this paper, it is important to recognise that an absence of reform is different from an absence of debate. After a thorough discussion of the issues, participants may conclude that any proposal likely to receive the necessary support would risk creating more problems than it would solve. Such a result should not be consid- ered a squandered opportunity because constitutional design always involves trade-offs — it is entirely appropriate for actors to decide that the status quo is, despite its flaws, superior to the alternatives. For this reason, my focus is on the nature of the deliberations that occur after the failure of a constitutional convention rather than merely the reforms that eventuate.

IV. Australia’s Constitutional Crisis of 1975

A. Two Questions

The events of 1975 provided an opportunity to revisit two fundamental ques- tions about Australia’s constitutional order, both of which deserved a fresh look given a number of changes in circumstances that had occurred since the Constitution’s enactment in 1901. First, what role should the Senate play in a parliamentary system that operates according to the principle of responsible government? As mentioned above, Australia’s federal government is a fusion of the American and British models of government. Similar to the United Kingdom, it operates in accordance with the principle of responsible govern- ment whereby the executive is drawn from, accountable to, and removable by the House of Representatives.¹⁹ Yet similar to the United States, the upper chamber, the Senate, is comprised of democratically elected members and invested with almost identical law-making powers to the lower chamber.²⁰

19. David Hamer, Can Responsible Government Survive in Australia? (Department of the Senate 2004) xvii. For the classic statements on the Westminster form of respons- ible government, see AB Keith, Responsible Government in the Dominions (Clarendon Press 1912); Anthony Harold Birch, Representative and Responsible Government: An Essay on the British Constitution (University of Toronto Press 1964).
20. There are some small differences between the legislative powers of two chambers, for example, proposed laws appropriating revenue or moneys, or imposing taxation, cannot originate in the Senate: Australian Constitution, s 53.

This fusion creates a tension. Given the Senate’s electoral credentials, it is far from clear that the Senate should yield to the House of Representatives in the event of a conflict between the two. This stands in contrast to many other parlia- mentary systems such as Canada, Ireland, New Zealand (prior to its abolition in 1950) and the United Kingdom where only the lower chamber is directly elected and thus has a strong claim to priority over the upper chamber in the event of a conflict between the two. However, it is also far from clear that the Senate should take actions – for example, blocking supply or preventing the govern- ment from implementing the policies which it was elected to implement – that are likely to lead to the defeat of a government and the formation of a new one, as these roles are the responsibility of the lower chamber in a system of respon- sible government. While there is an elaborate set of procedures for resolving deadlock between the two chambers involving mandatory waiting periods, double dissolution elections and joint sittings,²¹ they tell us little about which issues are appropriate for inter-chamber disagreement or how regularly and vigorously these disagreements should be pursued.
This tension has become more acute with the passage of time. The framers of the Australian Constitution envisaged that the Senate, like its counterpart in the United States, would be a mechanism for representing the interests of the states, especially the smaller states. The Senate is organised into state- based electorates with each state given an equal number of seats.²² Yet the for- mation of well-disciplined political parties at the federal level has effectively killed the Senate’s role as a voice for the states.²³ As the same political parties came to vie for control of both chambers, disagreements between the House of Representatives and Senate now tend to be partisan rather than federal in character. This tension reaches its zenith when the opposition party controls a majority of seats in the Senate because its Senators then wield a power that the principle of responsible government denies to its members in the House of Representatives, namely the ability to control the government’s legislative agenda. The opposition party can use its majority in the Senate to determine which laws the government can and cannot pass.
The second set of issues raised by the events of 1975 concern the role and identity of Australia’s head of state, the Governor-General. The Constitution vests this person, as the Queen’s representative, with a number of important functions, including the power to prorogue and dissolve Parliament and to appoint and dismiss members of the executive, including the Prime Minister.²⁴

21. ibid s 57.
22. The territories are also now represented in the Senate: Senate (Representation of Territories) Act 1973 (Cth).
23. The interests of the states are now primarily advanced through meetings of the fed- eral and state governments at forums such as the Council of Australian Governments, which was established in 1992, and through court challenges to the legality of actions taken by the federal government.
24. Australian Constitution, ss 5, 64.

A constitutional convention exists that the Governor-General, like the Queen in the United Kingdom, ordinarily exercises these powers in accordance with the advice of the Prime Minister.²⁵ This begs the questions of when, if ever, should the Governor-General depart from this convention and what should the Governor-General do when there are doubts about who should occupy the office of Prime Minister? While these questions arise in several countries, Australia’s system of government adds some peculiar twists, as the events of 1975 highlighted. The most striking is a product, once again, of the Senate’s legislative powers and electoral credentials.
In the event of a conflict between the House of Representatives and the Senate, there may a number of paths available for breaking the deadlock. The Governor-General may exercise his or her powers to call a double dissolution election, sending the members of both chambers to the polls,²⁶ but there is also the possibility of postponing invocation of this power if an ordinary election of one or both of the chambers is on the horizon — indeed, there may also be the possibility of bringing that election forward. When choosing from among these options, the Governor-General is placed in a difficult position. While con- vention suggests that he or she should act on the advice of the Prime Minister, the Prime Minister may attempt to push the boundaries of what is possible to obtain political advantage. Further, it is not beyond contention that the Prime Minister should prevail in a dispute with the Senate because, unlike other coun- tries in this position, the Constitution creates a Senate that is co-equal with the House of Representatives, holding almost identical legislative powers and elec- toral credentials. Thus, any intervention on the part of the Governor-General in these circumstances, whether to heed the advice of the Prime Minister or not, could be construed as a partisan intervention in favour of one political party or the other.
This issue has become increasingly vexed over the course of the 20th cen- tury because, as Australia acquired independence from the United Kingdom, it has left the government with a head of state that is appointed by, and a repre- sentative of, a foreign country. Australian independence was never reflected in the text of the Australian Constitution, but instead in a series of statutes such as the Statute of Westminster 1931 (UK) and the Australia Acts 1986 (Aus & UK). As a result, the provisions referring to the office of Governor-General remain unchanged. Section 2 of the Constitution continues to read that ‘[a] Governor- General appointed by the Queen shall be Her Majesty’s representative in the Commonwealth, and shall [hold office] during the Queen’s pleasure’. In acknowl- edgement of Australia’s independence, the office is now filled by an Australian rather than a Briton, but the electorate continues to have no direct say in the appointments process. The Queen is responsible for appointing and dismissing

25. As is the case with many other conventions, the existence and precise content of this convention is disputed: see, eg, Marshall (n 6) ch 3.
26. Australian Constitution, s 57.

the Governor-General and ordinarily exercises these powers on the advice of the Prime Minister of Australia.
The Governor-General is, therefore, an increasingly archaic aspect of Australia’s system of government. Prior to the events of 1975, however, there was limited interest in revisiting this issue. It generated only sporadic debate for practical reasons – these arrangements are embedded in a constitutional text that is difficult to amend – and, importantly, because it worked reasonably well. Apart from a dispute between the Australian and UK governments in 1930 over whether an Australian rather than a Briton could be appointed to the office, the identity and role of the Governor-General was generally not a cause of con- troversy and, moreover, had distinct advantages over the possible alternatives. The appointments process deprives the Governor-General of any electoral legitimacy, providing him or her with a strong reason to follow the advice of the Prime Minister and, as far as possible, to remain ‘above politics’.²⁷ This cre- ates constitutional stability by ensuring that there are not two offices capable of making claims to represent the people.²⁸ Replacing the Governor-General with a president could, if that person were elected, create the conditions for conflict between the Prime Minister and president. Thus, the issue increasingly merited detailed discussion not only because Australia’s relationship with the UK had changed, but also because there were strong arguments on either side. On the one hand, the pre-existing arrangements deny the people a direct say in the person appointed to serve as their head of state but, on the other hand, the Governor-General’s lack of electoral credentials contributes to harmonious, stable government.

B. The Events of 1975

After 23 years of government by the Liberal Party and Country Party (‘the Coalition’), the longest unbroken run in Australian history, Gough Whitlam led the Labor Party to a narrow victory in 1972, promising to implement a sweeping set of reforms. It did not, however, hold a majority of seats in the Senate, which had last faced the voters in 1970. The Coalition joined with the Democratic Labour Party (‘the DLP’) to prevent a number of the Labor Party’s legislative proposals from passing the Senate. Frustrated with the Senate’s obstinacy, Whitlam called a double dissolution election, forcing all members of the House of Representatives and Senate to the polls, after just one-and-a-half years in power in the hope of obtaining a majority in both chambers. The 1974 election returned the Labor Party to power in the House of Representatives, but left it two seats short of a majority in the Senate. Of the 60 seats in the Senate, the Labor Party and Coalition each held 29 with the remaining two held

27. Zelman Cowen, ‘The Office of Governor-General’ (1985) 114 Daedalus 127, 128.
28. Bruce Ackerman describes this approach to constitutional government as monist: see, eg, Bruce Ackerman, ‘Constitutional Politics/Constitutional Law’ (1989) 99 Yale Law Journal 453, 462–65.

by independents. In 1975, Whitlam appointed a Labor Party senator to the High Court and another Labor Party senator died in office, creating two casual vacan- cies. Under the Constitution, the legislature of the state from which the senator came is responsible for filling a casual vacancy.²⁹ A constitutional convention existed that the state legislature would appoint someone from the same polit- ical party as the former senator. In both instances, the state legislature broke with convention, giving the Coalition an outright majority in the Senate.³⁰
Amidst a series of scandals within the Labor Party, the Coalition used its majority to block supply in the Senate, leaving the Whitlam government unable to pass its budget. The Governor-General, Prime Minister and Leader of the Opposition discussed different options for breaking the deadlock (the Prime Minister preferred a half-Senate election while the Leader of the Opposition preferred a half-Senate and a House of Representatives election), but they could not come to an agreement. On 11 November 1975, with supply running out, the Governor-General, John Kerr, broke with convention by announcing that, even though the Prime Minister retained the confidence of the House of Representatives, he would not follow the Prime Minister’s advice and he dismissed Whitlam from office. He appointed the Leader of the Opposition, Malcolm Fraser, as Prime Minister on the condition that he would secure supply and advise the Governor-General to call a double dissolution election, which Fraser did. In the ensuing election, the Coalition was returned to power in the House of Representatives with a majority of seats in the Senate.

C. Opportunities Squandered, Opportunities Seized

The Governor-General’s departure from constitutional convention, precipitating the most significant constitutional crisis in Australia’s history, presented legal and political actors – as well as the public – with a rare opportunity to revisit the two fundamental questions about the country’s system of government mentioned above. The opportunity to address the first question about the Senate’s role in a system of responsible government was squandered, however, the opportunity to address the second question about the identity and role of the Governor- General was seized, prompting a national debate about whether Australia should become a republic.
While the actions of key individuals were the subject of intense scrutiny in the immediate aftermath of the crisis, outside of the academy there was little sustained public discussion about the tension the events revealed between the American (the Senate) and the British (responsible government) influences on

29. Australian Constitution, s 15.
30. In one instance, there was some uncertainty as to whether a breach had occurred because the state legislature appointed a member of the Labor Party, Albert Field, but the Party had rejected Field as a candidate because he had publicly said that he would not support the Whitlam government and was expelled from the Party, taking his seat in the Senate as an independent.

the Australian Constitution. Political actors predictably felt the need to take, and be seen to be taking, action to prevent a repetition of these events, but the measure that was the focus of discussion, and ultimately adopted, responded to the proximate rather than the ultimate cause. In 1977, the Constitution was amended to require a casual Senate vacancy to be filled by a person from the same political party as the former senator. While this amendment addressed the specific chain of events that gave the Coalition a majority in the Senate in 1975, it left the larger, more general questions about the Senate’s role untouched, such as: should the Senate be able to block supply, and how far should the Senate go in frustrating the government’s legislative agenda?³¹ Even within the academy, constitutional and political scholars have tended to overlook or underemphasise the issues surrounding the Senate’s compatibility with responsible government. There was, and continues to be, a tendency to describe the events of 1975 as a product of the particular personalities and fac- tual circumstances, sheeting home blame to Kerr’s decision to dismiss Whitlam, rather than as a moment of constitutional significance. As Brendan Lim writes,

The extraordinary events of the Whitlam years have so far assumed their place in constitutional history as an object of claim and counterclaim, but have come to be regarded as an aberration upon the tradition. They are not understood as a legitimate source of constitutional principle – not yet a part of the constitutional canon, demanding sustained reflection by successive generations who attempt to articulate constitutional meaning for their own time.³²

Constitutional scholars and government actors did, however, use the events of 1975 to initiate a sustained public debate on the second question regard- ing the appointment, tenure and identity of Australia’s head of state. The crisis spurred the creation of a strong republican movement advocating for consti- tutional change to replace the Queen and Governor-General with a president. The anger and emotion surrounding the dismissal prevented any rational dis- cussion in the years immediately after 1975. (Whitlam, for example, never spoke to Kerr again. Kerr resigned in 1977 and spent the rest of his life in Europe.) After these feelings subsided, however, members from both major political parties launched a national conversation on the subject. As Whitlam’s speechwriter Graham Freudenberg said, ‘the residual rage over the conduct of the Queen’s representative found a constructive outlet in the movement for the Australian Republic’.³³

31. There are some limited exceptions: see, eg, Minutes of Proceedings and Official Records of Debates of the Australian Constitutional Convention, Perth, 26–28 July 1978 (1978), 205 (Resolution 5).
    32. Brendan Lim, Crisis and the Canon: Australia’s Constitution after Whitlam (JSD Dissertation, Yale University, 2013) 4.
33. Graham Freudenberg, A Certain Grandeur: Gough Whitlam’s Life in Politics (revised edn, Penguin 2009) 463.

Debate began to gather momentum when Paul Keating, a leading advocate for republicanism, became Prime Minister in 1991 and influential bipartisan lobby groups were established in support of reform (the Australian Republican Movement) and in support of the status quo (Australians for Constitutional Monarchy) in 1991 and 1992 respectively. Keating established the Republican Advisory Committee in 1993, which began a wide ranging consultation pro- cess into options for change. The constitutional issues raised by the events of 1975 were at the forefront of discussion.³⁴ A central point of contention was, if Australia were to become a republic, the extent to which guidelines should be set down as to when the president could exercise his or her reserve powers, including dismissal of the Prime Minister.³⁵ When John Howard, an avowed mon- archist, became Prime Minister in 1996, he continued with the plans to hold a referendum on the issue, but steered the proposal towards an option that would minimise its chance of success. The model put to the people at a referendum in 1999 proposed to introduce a president appointed by Parliament rather than by direct election. While this would minimise the risk of the president claiming a popular mandate on behalf of the people to challenge the authority of the Prime Minister, it was also deeply unpopular with the public. It was perceived as an elitist model that deprived voters of a say in who should be their head of state, leading to its defeat at the referendum.³⁶ While this result put an end to dis- cussion for the time being, events occasionally breathe life back into the issue.³⁷

V. Patriation of the Canadian Constitution in 1982

A. Debating the Constitutional Amendment Procedure

The proposal of the Prime Minister, Pierre Trudeau, to break with convention in 1980 by proceeding to amend the Canadian Constitution without the con- sent of the provinces created an opportunity to recast the terms of debate about Canada’s constitutional amendment procedure and, more generally, the nature of its federal system. When the Canadian Constitution was drafted in the mid-19th century, the framers proceeded on the basis that the new federation would remain a part of the British Empire and did not, therefore, see any need to include a general provision for amending the document. It would be altered

34. See, eg, Report of the Republican Advisory Committee, An Australian Republic: The Options — Volume 1 — The Report (Commonwealth of Australia 1993) 28, 32, 88–116.

35. Ibid 88–116.

36. The proposal was defeated by a margin of 54.87% voting no to 45.13% voting yes nationwide and failed to obtain a majority voting yes in a single state.
37. See, eg, the decision of the Prime Minister, Tony Abbott, in 2014 to reinstate knights and dames to the Australian honours system and in 2015 to award a knighthood to the Duke of Edinburgh, Prince Philip: John Warhurst, ‘Abbott’s Knighthood Decision for Prince Philip Adds Fuel to the Republican Debate’ The Sydney Morning Herald (28 January 2015) <http://w w w.smh.com.au/comment/abbotts-knighthood-decision- for-prince-philip-adds-fuel-to-the-republican-debate-20150128-12zbwb.html> .

in the same way that it was enacted – by an Act of UK Parliament. The practice developed, beginning in 1895, that UK Parliament would amend the Canadian Constitution after receiving a request communicated by a joint address from the two Houses of federal Parliament in Ottawa.
An additional practice developed whereby Canada’s federal government would obtain the consent of the provinces before submitting a request to the UK for any proposed amendment that directly affected the powers of the prov- inces. Prior to 1982, there were five occasions where a proposed amendment met this criterion and in each instance the federal legislature obtained the unanimous consent of the provinces before proceeding.³⁸ Furthermore, there were at least three occasions where a proposed amendment was abandoned due to a lack of unanimous provincial consent.³⁹ There was, therefore, a strongly arguable case that a constitutional convention existed requiring the federal leg- islature to obtain the consent of the provinces prior to requesting a constitu- tional amendment that directly affected the provinces.
Practice consistent with this convention had stymied efforts to introduce a domestic amendment procedure that did not require the involvement of the UK — what came to be called ‘patriation’ of the Canadian Constitution. Conferences between the federal and provincial governments were held in 1927, 1931, 1935, 1950, 1960, 1964, 1971, 1978, 1979 and 1980 where the sub-
ject was discussed, but no action was taken because no consensus was reached on the amendment procedure.⁴⁰ In 1964, the conference of first ministers agreed on the‘Fulton-Favreau formula’, which would have required unanimous provin- cial consent for any amendment that would affect the property, powers, rights and privileges of the provinces.⁴¹ The proposal failed when the Quebec Premier, Jean Lesage, withdrew support in 1965. In 1971, the Prime Minister, Pierre Trudeau, proposed a set of amendments to the Constitution called the‘Victoria Charter’, which included an amendment provision that would have given the two largest provinces (Ontario and Quebec) and a sufficiently large coalition of either the Western or Atlantic provinces a veto over future amendments.⁴² Eight of the ten provinces agreed to the Victoria Charter, but it was abandoned after the Quebec Premier, Robert Bourassa, announced that the province would not consent due to concerns about provisions unrelated to the amendment pro- cedure. Debate on patriation gradually became circular and stale. Discussions

38. Reference Re: Resolution to Amend the Constitution [1981] 1 SCR 753, 891. However, there have been four instances where‘one or two provinces protested that federal-provincial consultations should have taken place prior to action by Parliament’: Favreau, The Amendment of the Constitution of Canada (n 13) 16.
39. Reference Re: Resolution to Amend the Constitution [1981] 1 SCR 753, 893–94. Two of them are discussed below: see text following n 40.
40. Ibid 904.
41. Anne F Bayefsky, Canada’s Constitution Act 1982 & Amendments: A Documentary History — Volume 1 (McGraw-Hill Ryerson 1989) 16–21.

42. Ibid 214–21.




would focus on proposals most likely to succeed in securing the consent of all the provinces, and especially Quebec, thus narrowing the scope of debate. Yet each proposal ultimately failed due to a lack of consent from all the provinces.

B. The Events of 1980–82

Trudeau initiated another attempt to amend the Constitution in 1980. However, the talks once again failed. His attempt to bundle together a number of dis- parate reforms led the provinces to take different positions on different issues and to assign different levels of importance to them, undermining efforts to find points of compromise.⁴³ In the past, this impasse would have marked the end of negotiations and led to the issue being taken off the table for the time being. However, Trudeau was determined to achieve constitutional reform and flagged the possibility of proceeding in violation of the purported constitu- tional convention. He announced that the federal government was willing to submit a request for amendment to UK Parliament irrespective of the consent of the provinces. He announced an interim, two-year period during which the federal and provincial governments would seek to reach unanimous agree- ment on an amendment procedure. If they did not reach unanimous agree- ment but eight or more of the provinces agreed, representing at least 80% of the total population, then that agreement and a procedure similar to the one found in the Victoria Charter would be put to the people at a referendum. If neither of these two scenarios eventuated, an amendment procedure similar to the one found in the Victoria Charter would be unilaterally pursued by the federal government.
Eight of the ten provinces opposed Trudeau’s decision, partly due to the unilateral nature of the federal government’s action and partly due to growing opposition to the Victoria Charter formula.⁴⁴ They decided to use a combination of political and legal avenues to pursue their opposition,⁴⁵ with the latter course involving challenges being simultaneously launched in the courts of three provinces. These decisions were eventually appealed to the Supreme Court. In a landmark decision,⁴⁶ a majority of the Court found that the federal government had the legal authority to submit a request to UK Parliament without provincial consent. However, a differently constituted majority found that a constitutional convention existed requiring the federal government to obtain ‘a substantial

43. For an overview of discussions, see Edward McWhinney, Canada and the Constitution 1979–82 (University of Toronto Press 1982) 43–45.
44. D Marc Kilgour and Terrence J Levesque, The Canadian Constitutional Amending Formula: Bargaining in the Past and the Future (Martinus Nijhoff 1984) 469.
45. The political opposition involved lobbying members of UK Parliament to block or delay the passage of legislation formally enacting the constitutional changes: McWhinney, Canada and the Constitution 1979–82 (n 43) 72–74.
46. Reference Re: Resolution to Amend the Constitution [1981] 1 SCR 753.

degree of provincial consent’⁴⁷ for an amendment that would directly affect the provinces. Importantly, the Court refused to provide further details as to what it meant by a ‘substantial degree’, which meant that the parties could not defin- itively rely on the judgment to say that a particular level of provincial consent would or would not comply with the convention.
Trudeau’s proposal to act unilaterally and the ensuing Supreme Court deci- sion provided an opportunity to recast the terms of debate between the federal and provincial governments. It changed the momentum by shifting the direc- tion of debate from one in favour of the status quo to one in favour of reform. The federal government’s determination not to be bound by the constitutional convention, and the Court’s confirmation that it was legally entitled to proceed unilaterally, left the provinces, especially Quebec, in a position where recalci- trance may result in a worse outcome than engagement. Yet the fierce provincial opposition to the proposal to act unilaterally, combined with the Court’s confir- mation of a constitutional convention requiring substantial provincial consent and the possibility that some members of UK Parliament might seek to obstruct the federal government’s plan if it proceeded without provincial consent,⁴⁸ provided Trudeau with sufficiently strong reasons to go back to the negotiat- ing table and seek a compromise with the provinces. An additional incentive for Trudeau to avoid unilateralism was the prospect that Quebec might seek to leave Canada, which was a distinct possibility after a referendum held in Quebec in 1980 asked whether the province should pursue a path towards sovereignty. Although the proposal was defeated, the event demonstrated the seriousness of Quebec’s nationalist movement and that unilateral action on the part of the federal government might serve as a catalyst for further steps towards inde- pendence. Thus, even though Trudeau, the person who broke with convention, stood to gain the most from that break, it was nevertheless a step fraught with peril for him, the federal government and, indeed, the entire country.
The threat to break with convention also had implications for debates regarding the future direction of Canada’s federal system. As the likelihood of reform substantially increased, political actors were confronted with an impending decision on the respective capacities that the federal and provincial governments would have in the future to shape Canada’s constitutional system through the amendment procedure. The provinces would have significantly more power to protect their interests if they were given the ability to veto future constitutional amendments, producing a more stable, less dynamic federal sys- tem than if the federal government could proceed with the consent of only some (or perhaps even none) of the provinces. This was well recognised. When the Fulton-Favreau formula, which would have required unanimous provincial consent to amend some parts of the Constitution, was under development and discussion in the 1960s, Saskatchewan was a vocal opponent, arguing that such



47. Ibid 904–5.

48. McWhinney, Canada and the Constitution 1979–82 (n 43) 90–91.

a rigid amendment procedure would stunt the growth of Canada by allowing lone holdouts to block changes viewed as beneficial by all the other govern- ments.⁴⁹ The threat to break with convention altered the calculus because, on the one hand, it placed the proponents of reform in a position to adopt a pro- cedure that did not give Quebec a veto over future amendments yet, on the other hand, risked exacerbating nationalist sentiments in Quebec if its voice was ignored or minimised during the negotiating process.

C. An Opportunity Seized, but with a Price

After the Court’s ruling, Trudeau wasted little time, convening another first min- isters’ conference to see if an agreement was now possible with the threat of unilateral action looming. Eight provinces began the conference opposed to the federal government’s plan, but seven of them changed their position follow- ing a last-minute round of negotiations that excluded Quebec.⁵⁰ They agreed to a general amendment procedure that would require future amendments to obtain the approval of at least two-thirds of the provinces having at least 50% of the population and that would not give any single province the power to veto future constitutional amendments.⁵¹ (There are exceptions to this general procedure. Unanimous provincial consent is, for example, required to amend the amendment procedure.) With the approval of nine of the ten provinces, the package of reforms was sent to UK Parliament and enacted.
While the threat to break with convention presented an opportunity to recast the terms of debate and achieve constitutional change, an opportunity that the federal government seized, it came at a high price. Quebec stridently denounced the outcome. The Quebec National Assembly unanimously passed a resolution condemning the settlement; a legal challenge was launched in the Supreme Court⁵²; and the notwithstanding mechanism contained within the Charter of Rights and Freedoms was invoked in a blanket fashion, with Quebec repealing its entire body of legislation and re-enacting it with notwithstand- ing declarations.⁵³ This hostile situation led the federal and provincial govern- ments, after Trudeau was no longer Prime Minister and Levesque was no longer Premier of Quebec, to seek an agreement that would lead Quebec to support the Constitution and the package of reforms. The Meech Lake Accord proposed

49. See generally BL Strayer, ‘Saskatchewan and the Amendment of the Canadian Constitution’ (1966–67) 12 McGill Law Journal 443.
50. See generally Guy Laforest and Rosalie Readman, ‘More Distress than Enchantment: The Constitutional Negotiations of November 1981 as Seen from Quebec’ in Lois Harder and Steve Patten, Patriation and Its Consequences: Constitution Making in Canada (UBC Press 2015) 159.
51. Canadian Constitution Act 1982, s 38(1).
52. The challenge failed: Reference Re: Amendment to the Canadian Constitution [1982] 2 SCR 793.
53. The Supreme Court upheld this use of the notwithstanding mechanism in Ford v Quebec (Attorney General) [1988] 2 SCR 712.

a number of changes to the Constitution to secure Quebec’s assent, includ- ing an amendment procedure that would give a veto to all provinces, essen- tially codifying the pre-1982 constitutional convention. After the Accord was sent to the provinces for approval, support began to unravel and it was never approved. Another attempt was made in 1992. The resulting package of pro- posed reforms, the Charlottetown Accord, would have increased the number of items in the Constitution requiring unanimous provincial consent to amend. It was defeated in a national referendum by a margin of 55% to 45%. Its failure added fuel to nationalist sentiments in Quebec, culminating in a referendum on independence in 1995, which was defeated by the narrowest of margins: 50.7% to 49.3%. Canada thus endured two decades of extensive constitutional uncer- tainty as a result of having seized the opportunity presented in the early 1980s.

VI. Conclusion

While there are, as many scholars attest, sound reasons for expressing quiet concern about constitutional conventions due to the uncertainty that sur- rounds their content and enforceability, it is, I contend, important to recognise the reasons for cautious optimism too. Those uncertainties can spur constitu- tional deliberation because they mean that, when a convention fails, actors must directly engage with various dimensions of the convention (its scope, consequences and rationale), will often need to work through multiple options for reform as the simplest solution (codification of the convention) might not be feasible, and may not be able to rely on the courts to resolve the issue for them (even if the courts are involved to some extent). As a result, actors are forced to confront the issue and engage with each other to find a solution. We should not welcome the failure of constitutional conventions given that they can throw a country into political crisis and generate enduring instability about fundamental features of government. However, we should also be attune to the opportunities they present for deliberation on constitutional issues that do not get the attention they deserve or that suffer from stale frames of reference. For countries that have highly stable constitutional orders where opportunities for reflection are rare, these deliberative moments are vital points of reconsider- ation and renewal that should be seized.

SCOTT STEPHENSON*







* Lecturer, Melbourne Law School, The University of Melbourne. Thanks to the parti- cipants of the Symposium on Conventional Constitutionalism, Trinity College Dublin, 15 November 2014 for their helpful comments and suggestions.