How to Govern in Coalition or Minority: The Case of New Zealand

Article 9 / 13 , Vol 36 No 4 (Winter)

How to Govern in Coalition or Minority: The Case of New Zealand

New Zealand switched electoral systems from single member plurality to mixed member proportionality for the 1996 election. The country’s leadership was well aware that this change would mean that no one political party would have a majority of seats in the legislature, so extensive study was undertaken in advance with respect to coalition and minority governments. While this advance work held the public service in good stead, the political parties failed to respond adequately to the new governing dynamics. Even with the leadership of a former senior jurist as governor general, it would take until Y2K for the political elites to learn how to operate within the new paradigm. The procedural improvements made by New Zealand in this period have most recently informed improvements to parliamentary government in the United Kingdom and Australia. This paper examines these and other lessons that New Zealand may offer Canada.

Canada, along with the United Kingdom1, Australia2. and New Zealand share the ‘Westminster-model’, so named because this design has been inherited from that used for the British at the Palace of Westminster. Also called ‘responsible parliamentary government’, a label that emerged here in Canada, it is a parliamentary system whereby the people elect representatives to a legislature and it, in turn, chooses a government. The process is guided by a set of unwritten constitutional conventions. And while these conventions offer specific guidance as to by whom and how decisions should be made, when it comes to the ‘reserve powers’ of the monarch or her governor general – dissolving parliament, proroguing a session and choosing or dismissing a prime minister – they have begun to operationalize differently in each of these countries.

The reason for the deviation it two-fold: First, the electoral landscape has changed in each of these countries from what had previously been a majoritarian norm. This norm was created by single member plurality voting which in most countries delivers a majority of seats in the legislature to a single political party.3 Even Australia, which had moved away from the SMP electoral system in 1919, was able to maintain majoritarian politics for the longest time through a semi-permanent coalition of two political parties on the right. But recently, beginning with New Zealand, each of these countries has seen its legislatures divided by multiple political parties.

Second, there has been a shift in political culture. Notions such as the need for government to implement policies that are supported by the majority of the legislature (and by extension the majority of the population), for fairness to minority political parties, for greater openness and accountability in government decision making and to increase civility in public life have created pressures in a number of Anglo countries to revisit the electoral system and to clarify the constitutional conventions that govern the Westminster-model.

In response to public demands for fairness to minority political parties and the voters who support them, New Zealand appointed a Royal Commission on the Electoral System which recommended a shift to ‘mixed member proportionality’ in 1986. Over the next few elections, the idea of holding a referendum on a change to the electoral system became a key election issue and, in 1992, the government was forced to follow through on its campaign promise, though it announced that the referendum would be non-binding. When 84 percent of the voters expressed the desire to change the system and 71 percent indicated that MMP was the preferred alternative, the government tried to backpedal by holding a second referendum. This one would be binding and held during the general election the following year, pitting SMP directly against MMP. In spite of a heavily funded campaign for SMP endorsed by many high profile political and business elites, MMP was chosen 54 to 46. Parliament then adopted MMP beginning with the 1996 general election.

What is particularly important is that the New Zealand legislature adopted this change with the full understanding that it would mean the end of any one political party having majority control of the country’s parliament. This would mean either coalition governments, which is the norm in most parliamentary democracies that have electoral systems where no political party wins a majority, or minority governments where negotiations for support on financial matters (supply) and confidence questions are undertaken following the election to ensure that the government has the support of parliament before being sworn into office (the norm in most minority government situations outside of Canada).

The first thing New Zealand’s elites did was to undertake comparative research to prepare for the transition. This began in the academy as would be expected. But it was quickly taken over by the different branches of government, including the governor general, parliament and the public service. This included trips overseas and commissioned papers on questions like government formation and constitutional conventions.

Obviously the first lesson to be taken from New Zealand is the importance of comparative research to prepare for all eventualities. But it also raises the important question of, even with foresight and all this research and planning, what did New Zealand do well and what did it do poorly?

What New Zealand did poorly was at the political level, where in spite of planning it took time for politicians to master the art of government formation and government administration. The four things that it did well were to:

(i) choose governors general who would be able to interpret the constitutional conventions and apply them fairly and, in the case of the first GG, was confident enough to break with tradition and impartially help the media and the public understand the constitution and the process,

(ii) make plans at the bureaucratic level for the challenges of uncertainty in a system of government that had previously been efficiently dichotomous in terms of political leadership,

(iii) release publically a cabinet manual so that all political actors could inform themselves about these conventions and improve upon this document in response to the unforeseen challenges that the first divided parliament presented for coalition governance, and

(iv) make clear from the start the rules surrounding ‘caretaker governments’ to instill confidence in financial markets that there was still a government in place that could deal with a crisis while at the same time instilling confidence in the political leadership that they would not be hamstrung by any decision of this former government while they explored alternative government configurations.

These five items will be examined in this order in the following five sections.

Government Formation

Even though the political parties went into the 1996 general election knowing no party would win a majority of seats and had acknowledged that a coalition government was a likely and legitimate outcome, the learned behaviour of the politicians had been acquired in majoritarian politics and they were inexperienced on how to negotiate and how to build and maintain trust which is essential for stability in a government.

The decision in New Zealand was to let MMP occur under the existing Westminster-model constitutional conventions. This was a conscious choice. Responsible parliamentary government was predicated on the executive branch’s accountability to parliament and this historically had led to inter-party negotiation in New Zealand when no party had a majority of seats. But these experiences had occurred in the context of an expected majoritarian government next time around.

Under section 19 of the Constitution Act 1986, the New Zealand Parliament must meet within six weeks of the return of the writs for a general election; and section 17 says the term of Parliament ends three years after the return of the writs unless Parliament is dissolved earlier by the governor general. The short parliamentary term means that ‘snap’ elections and prorogation are not issues in New Zealand.

In the first election under MPP in 1996 election, the National Party won 44 seats, Labour 37, New Zealand First 17, Alliance 13, ACT 7 and United New Zealand 1. New Zealand First began negotiations with the two largest parties. The majoritarian habits and the lack of experience with ordered bargaining resulted in a drawn out and uncertain bargaining process.4 Boston and Church have characterized these negotiations as NZ First “holding the country to ransom”.5 It took two months to negotiate a coalition government.

The immediate reaction to this long and messy government formation was academics and politicians revisited their earlier conclusion about honouring the existing constitutional conventions of the Westminster-model. Among the many recommendations was that the governor general appoint the leader of one political party following the election who is most likely to be able to form a government or, in the alternative (so as to keep the GG above the political fray), have the speaker of parliament choose the party leader to take the first kick at the can.6 In the end, no changes were made to the constitutional conventions and the reason for this can be attributed to the governor general who repeatedly reassured New Zealanders of the soundness of the rules.

Forming a coalition government is one thing. Governing in coalition is another, and it requires building trust, understanding the rules and effective dispute resolution mechanisms. The biggest trust challenge is with the minority governing partner. So, not surprisingly, the first public problem in the coalition emerged, when the Associate Minister of Health (an MP from NZ First) had to be fired by the PM after his continued fighting with the Minister of Health (an MP from National) and his public criticism of the coalition just over halfway through the first year.7

In New Zealand the party leader is chosen and removed by the parliamentary caucus and this happened in the National Party in December 1997, just under a year after the PM had been sworn-in, resulting in a change in PM. By August of 1998, the NZ First leader was fired as Deputy PM and Treasurer (a post created for him as part of the coalition agreement) after a very public dispute over the privatization of the Wellington International Airport. This ended the coalition with NZ First, which caused an exodus from the party of MPs who had been ministers and some backbenchers. Most of these formed a new party called Mäuri Pacific and, along with one Alliance MP, they formed a new coalition with National which was able to govern until the 1999 election.

There was no systemic reason for this breakdown within and between the coalition partners. As previously noted, coalitions require trust building and dispute resolution mechanisms and the cabinet system lent itself to both things. In New Zealand they have (i) two tiers of ministers – minister and associate minister – and (ii) a long standing practice of appointing ministers who stay outside cabinet. These type of mechanisms are used in a number of countries where parliaments are divided to create stability. Associate ministers can be appointed for specific issues that are of concern to a coalition partner, to allow for departmental input where a portfolio is held by a different coalition partner or, where a political party only has a few members, create a more generalist mandate so the associate minister can have input over a number of government departments.

So why did the coalition government breakdown in this first MMP parliament? There was unfamiliarity with the constitutional conventions and ministerial/cabinet responsibility, which will be addressed more fully in the section on the cabinet manual. But the real problem lay in the inexperience among the political class with coalition governments. This was exacerbated by an almost panicked response to shifting public opinion and a lack of understanding about what the shifts meant for each party in the coalition in the next general election. The senior coalition partner did little to build and maintain trust and the junior partner was insecure about what it would mean to face an electorate and run on the record of a coalition government where it was the junior partner.

The fact that the coalition negotiations had been done in secret over two months was a factor. The leadership of New Zealand First took short-term satisfaction in being courted by both political parties but when it finally began in earnest to negotiate with National then the concessions it achieved were not readily understood by its membership and the public. Ironically, all the flaws in government formation New Zealand experienced had been discussed in the plethora of comparative research that parliament, the New Zealand government and academia had compiled in the run-up to 1996.

Learning from the previous government’s mistakes, the second coalition government formed in 1999 was installed after only 10 days of negotiation, comparable to the average formation period under the previous electoral regime, and the issues being negotiated and concessions made were much more public and limited (to process over policy).8 This was a coalition between Labour and the Alliance Party with negotiated support from the Greens for confidence and supply. Following the 2002 election, Labour formed a coalition with the Progressive Party with negotiated support from the Greens and United Future. After the 2005 election, Labour formed a coalition with the Progressive Party with negotiated support from NZ First and United Future and a signed agreement from the Greens that they would abstain on confidence and supply votes (the Mäori Party also abstained on these votes but had no formal agreement). In 2008, National formed a minority government with negotiated support for confidence and supply matters with ACT, United Future and Mäori parties and this was continued after the 2011 election.

To briefly illustrate how ministerial posts can be used to obtain support from other political parties and build trust (and this is just one small example from a large number of possible mechanisms), there are currently two ministers who are not members of the cabinet appointed from one of the parties that have agreed to support the government on confidence and supply votes. Table 1 shows that these two ministers were each assigned a series of delegated authorities as a mechanism to allow them to have oversight of the government inside the executive branch and to facilitate their involvement in specific files across portfolios. This helps to engender confidence that the issues important to the smaller political parties supporting the government will be followed through, as promised in the written agreement.

Table 1
NZ Ministers outside Cabinet
Dr. Pita Sharples

Co-leader of Mäori Party

Minister of Mäori Affairs

Associate Minister of Corrections

Associate Minister of Education

Tariana Turia

Mäori Party MP

Minister for Disability Issues

Minister for Whanau Ora

Associate Minister of Health

Associate Minister for Social Development and Employment

Associate Minister of Housing

Associate Minister for Tertiary Education, Skills and Employment (relating specifically to the Employment area)

There continues to be political disagreements in New Zealand, but that is the normal cut and thrust of legislative and cabinet debate. New Zealanders have mastered government formation and management under the new paradigm of divided parliaments and power sharing. Absent is the brinksmanship and vitriolic discourse that has plagued the Canadian Parliament for some time. Aucoin et al. have suggested that this is an artifact of Canada’s minority governments.9 They are correct in so far as it is an artifact of minority governments where there is no formal negotiated agreement for legislative support on confidence and supply. The evidence from New Zealand, where there is no incentive to go to the polls early and where the first coalition government was by any objective means a failure at the ministerial-level, is that the bad behaviour seen in Canada is an artifact of majoritarian politics (not minority government per se).

Governors General

Knowing that in a divided parliament the governor general’s reserve powers would be a factor in government formation and for mediating relations between the executive and legislative branches, and wanting to ensure that the office of the governor general and thus the monarch was kept above the fray, the prime minister asked the Queen to appoint a former New Zealand Court of Appeal Judge as governor general in advance of the 1996 election.

Upon receiving his appointment, Sir Michael Hardie-Boys immediately began his own investigation on the constitutional conventions surrounding what are called in New Zealand (like Canada) the ‘reserve powers’ as they remained in the hands of the monarch in an era where most royal prerogatives were being turned over to the executive branch. These mediate relations between the two branches and include the power of dissolution, prorogation and the power to appoint and dismiss a prime minister.

In addition to reviewing his constitutional texts, Hardie-Boys opted to undertake his own comparative examination of how heads of states are involved in government formation, beginning with a visit to Ireland and Denmark immediately following his appointment. This complimented the comparative research which the public servants had commissioned (discussed below).

Governors general rarely make public their intended action and, in many jurisdictions, do not even provide information directly to the public about the decision made after the fact (leaving it to the prime minister who, as Canadians know, will often spin the decision in the current context for partisan gain). Given the major changes about to occur in New Zealand, Governor General Hardie-Boys decided to launch a public campaign to educate New Zealanders (and the political elite) about the role of the governor general in government formation. This began with a widely publicized speech, followed by media interviews to clarify key points, and ended with his participation in a documentary, televised shortly before the election. Hardie-Boys explains the role of this public education campaign:

The aim was to ensure, so far as possible, that the principles and processes for moving from the election to the formation and appointment of a new government were clear and understood by a sufficient number, so that the focus of public attention could be where it belonged – on the political actors who would be required to negotiate and work together to reach a political resolution.10

The key points Hardie-Boys made during these pre-election interviews were:

  • Government formation is a political decision to be arrived at by the elected politicians.
  • The governor general must ascertain where the support of the House lies and that means, in an unclear situation, communicating with the leaders of all of the parties represented in parliament.
  • Once the political parties made public their intention to form a government the governor general may have to talk with the party leaders to obtain sufficient information to ensure he is appointing a PM (if that is what is required) who has the support of the House.
  • During negotiations the incumbent PM remains in office but is only governing in accordance with the ‘caretaker convention’.

And the day after the election Hardie-Boys issued a press statement reiterating the key points about the process for government formation.

The Electoral Commission, which had its own public campaign to educate New Zealanders about the new MMP electoral system, also provided information about the role of the governor general, the reserve powers and the concept of a caretaker government

A review of the media reports during and immediately following the election suggests that Hardie-Boys and the Electoral Commission were successful in their messaging. The New Zealand public was informed repeatedly of how the process would unfold and media coverage was entirely focused on the inter-party negotiations occurring in parliament without reference to what the governor general might do in certain circumstances, thus insulating the office from what turned out to be a protracted partisan negotiation.

While Hardie-Boys had publically asserted his constitutional right to consult with all political leaders either during the negotiations or upon their conclusion so as to confirm the individual he invited to form a government did, in fact, have the confidence of the House, behind the scenes he assigned the Clerk of the Executive Council this responsibility. He also authorized her, as his representative, to assist the parties in their negotiations concerning the logistics of government formation.

It was noted above that this first government formation took two months to negotiate. It should also be noted that the governor general had in his public comments suggested that it was better to take the time to negotiate a well-considered government (there was, after all, a caretaker government still in place) and he had informed the public that the limit to the negotiations was eight weeks (the constitutional requirement noted above meant that parliament had to meet by December 13). Agreement was reached on December 10.

In the remaining years until his retirement from the post in 2001, Governor General Hardie-Boys continued to periodically talk publically about the role of the governor general, including his decisions in 1996, and, as noted above, by 1999 the political parties had begun to learn the art of government formation.

Hardie-Boys’ appointment was followed by Dame Silvia Cartwright, a former High Court Judge, and in 2006, by Sir Anand Satyanaud, a former District Court Judge and the country’s Ombudsman. While these jurists have not had to deal with the challenges or be as proactive as Hardie-Boys, this tradition of selecting from the judiciary people with the necessary skill-set to manage the reserve powers was important during the two decades following the move to MMP and divided parliaments.

Since 2011, the governor general has been Sir Jerry Mateparae, the second GG of Mäori descent and the first Mäori to reach the rank of Chief of the New Zealand Defence Staff. While not a jurist, the symbolic representation his appointment has for the Mäori community and the GG’s role as commander-in-chief commend this appointment now that the need for a jurist who can oversee government formation and administration has lessened. Given his credentials, and the clarification of the constitutional conventions by Hardie-Boys, there is little doubt he can navigate the reserve powers if called upon to do so and not be a docile handmaiden to the prime minister (which a number of PMs in Canada have insisted ‘their’ governors general be).

Bureaucratic Planning

Where advance planning made the biggest difference was in the public sector. It was identified early on that public servants would be faced with periods of uncertainty as they waited for a new government to be negotiated and, once formed, there would be the additional challenge of accommodating inter-party politics within the ministerial ranks of the executive branch.11 As noted above, the public servants launched their own review of how coalition governments are formed and operate in other developed democracies.12 This included sending public service delegations overseas.13

Additionally, and I might even argue more importantly, the public service (with the permission of its political masters) commissioned academic research and expert advice. Detailed analyses were made of coalition arrangements in various European countries, and these were then applied to different models of governance for New Zealand.14

One of the lessons public servants learned from their comparative research is that government formation can and will inadvertently involve public servants in government formation negotiations. Proper planning can keep them out of partisan discussions and ensure their neutrality (as they will be required to work with whatever government emerges).

Rules were established in advance of the 1996 election to allow for political parties to obtain information to support negotiations over a policy programme while ensuring the neutrality of the public service. These included:

  • Public servants could only provide information to political parties when requested and authorised by the prime minister (who was not to be shown the response to any request unless it came from his own party);
  • If ministers in the caretaker government wanted information to use in the negotiations, they had to request this through the prime minister and not approach their own department directly;
  • All requests for information and any resulting written briefings were to be channeled through a committee of senior officials, including the Cabinet Secretary;
  • Public servants were only to comment on the practical implications of any policy proposal and not its merits; and
  • No public service input would be provided for the drafting of a coalition agreement (which was a matter solely for the political parties).

Part of the reason for the protracted negotiations in 1996 was that New Zealand First wanted a detailed policy agreement, so it made a large number of requests costing various policy proposals.15 In reflecting on the process, the public service concluded that filtering requests and responses slowed down the process. Political parties needed responses within a few days and as public servants needed those days to do their analysis, delays in communication meant delays in negotiations. There was also concern that the filtering of the responses resulted in briefings that were of only marginal help to the negotiators. A subsequent review of the arrangements suggested that more flexibility might be beneficial, and that direct contact between the negotiating parties and public servants might help reduce some of the misunderstandings and confusion that arise when correspondence is limited to written documents.

The guidelines produced prior to the 1999 election differed little from the previous rules, except in allowing for face-to-face meetings of officials and party negotiators once it was clear that the parties concerned were likely to form a government. Direct meetings could only be held prior to this stage when a written request for information was unclear. In those instances, a meeting to resolve the issue would be attended by the relevant deputy minister (what they call permanent secretary) and by officials from the Department of the Prime Minister and Cabinet (our Privy Council Office) and the State Services Commission (our Public Service Commission), to ensure the impartiality of the public service was maintained.

The system, as noted above, worked much better in 1999, where it took only 10 days to negotiate an agreement; though it should be noted that there were fewer demands on public servants in this instance as the negotiations between Labour and Alliance focussed on procedures and not policy.

Cabinet Manual

There has been a comprehensive cabinet manual in New Zealand since 1979. It was initially a restricted document with distribution confined to the Cabinet Office, ministers and senior officials. In 1991 it was decided to make it available to all public servants as a loose-leaf publication. In anticipation of the 1996 transition to negotiated coalition or minority government, it was decided to redesign the document and make it publicly available so that the everyone, especially the political parties in parliament, would have a better understanding of the constitutional rules, conventions and processes. By 1998 it was available online to the public and the world. New Zealand became the first to make this kind of information on the internal operations of government, and the rules that guide and constrain it, publicly available.

It is not a codification of the unwritten portions of the constitution. It is not a legal document. And it is not justiciable. It is an internal document to cabinet itself. As such, it is adopted at the beginning of each government. New editions are authorized by the prime minister and then drafted by the Cabinet Office and submitted to peer review by senior officials in Crown Law, Department of Justice, State Services Commission and the Treasury. Specific chapters are sent to officials concerned with its subject matter, such as the Ombudsman, Privacy Commissioner and Clerk of Parliament.

Over time the Cabinet Manual has become a document of best practices for government decision making. It has removed uncertainty surrounding the procedures and practices of the cabinet, but most importantly it has become a useful tool in government formation, and in defining ministerial responsibility and collective responsibility; and not surprisingly it has become a useful tool to the media on those same issues, thus preventing misunderstanding and misinformation.

Rebecca Kitteridge, former Secretary to New Zealand Cabinet, has compared it to a dictionary: “authoritative, but essentially recording the current state of the constitutional and administrative language.”16 Like a dictionary it lags behind institutional developments; just as words are not included in a dictionary until they are part of the popular lexicon, constitutional conventions are not put in the cabinet manual until they have been firmly established as such:

“The key point is that although amendments to the Manual may reflect and promulgate change, they do not, in themselves, effect change. Change is effected by new legislation, or Cabinet minutes, or judicial decisions, or amendments to the Standing Orders. Even rules on the processes of executive government, which may not be recorded anywhere except in the Manual, are approved by Cabinet at the time the Manual is issued. Their authority derives from Cabinet.

“The fact that the Manual cannot, by itself, effect change is even more significant in respect of those provisions that articulate elements of the constitution. Clearly the constitutional conventions exist independently of the Manual, although they are authoritatively expressed there. So, for example, changing the provisions of the Manual relating to the constitutional powers of the Prime Minister, in the absence of separate constitutional developments, will not have any effect on the conventions themselves.”17

To continue with Secretary Kitteridge’s dictionary analogy, the editors of the Oxford English Dictionary use pigeonholes in their editorial offices (or ‘Scriptorium’) to file suggestions (or ‘slips’) from contributors. Once there are a sufficient number of slips, and subject to a consensus as to their deservedness for inclusion, a new edition will be published with the new words. In the same manner constitutional and institutional changes will make their way into the Cabinet Manual from officials’ and academics’ suggestions. When there is a sufficient number the Cabinet Office proposes a new edition to the PM.

The first revision following MMP was the 2001 edition and the changes are significant. First, the new Cabinet Manual was influenced greatly by the speeches of Governor General Hardie-Boys (in fact chapter 4 is largely a compilation of the constitutional conventions he identified that mediate government formation and political crises), and thus it is a much more solid reflection of responsible parliamentary government than what was put out in 1996 based on the internal review of cabinet governance by the executive branch.

Another change made in 2001 was that much of the detailed procedural guidance about Cabinet and Cabinet committee processes was removed from the cabinet manual and placed in a new Cabinet Office Step by Step Guide. It was only officially called the Cabinet Manual at this time. Previously it was called the Cabinet Office Manual, but the name change reflects its transformation from a book of procedures used by the Cabinet Office to a book on principles of executive government that guide cabinet and each minister including the PM.

The New Zealand Cabinet Manual has become useful for resolving disputes. For example, a section on ministers’ statutory powers and functions in the collective cabinet context has been added. It makes clear that while individual ministers take particular actions or decisions, the ministers do so within the framework of cabinet collective responsibility. If the decision or action would affect the collective interest of the government, the minister should not take the relevant action or decision without consulting relevant colleagues at an early stage and submitting a paper to cabinet. This section settled a dispute between the Cabinet Office, which held the view that a minister can generally consult with whomever the minister pleases before reaching a decision and should if there is an impact on other departments or on the government as a whole, and departmental officials who had taken the view that the minister can act autonomously in areas where he has statutory power and that it may even be inappropriate to discuss matters in cabinet as that could invite judicial review.

Not surprising, a number of changes to the 2001 version reflect problems that arose during New Zealand’s first post-MMP coalition government. For example, the Cabinet Manual now states that the portfolio minister always retains overall control of the portfolio, and that the associate minister only has delegated authority. This clarification was necessary due to the number of conflicts that arose between portfolio ministers and associate ministers when they came from different political parties (that conflict also arises in Canada where they come from the same party, such as at National Defence and Foreign Affairs, but party solidarity tends to keep these differences between the two ministers and easily resolved by the PM).

Similarly, paragraph 2.8 in the 2001 manual states “As the chair of Cabinet, the Prime Minister approves the agenda, leads the meetings and is the final arbiter of Cabinet procedure.”18 As noted above, the coalition ended in 1998 after a dispute over the privatization of Wellington International Airport and, while cabinet usually works on consensus, in this coalition government it became impossible; cabinet procedures, such as the quorum for the cabinet meeting where privatization was approved, became points of contention in the absence of unanimity. This clarification is intended to prevent future disputes over cabinet procedures.

Whether or not there is a lesson for Canada in this publication is dependent on the willingness of the government of the day to use a process, internal but impartial like New Zealand or public and multi-party like the United Kingdom, that would ensure the end product contains best practices and is designed to improve the functioning of responsible parliamentary government. As I have written before in this publication and told the current government, to create a manual that is designed to distort conventions in favour of the executive branch will result in a manual that will be without credibility or, worse, do damage to Canada’s institutions which are already suffering from declining public confidence due to misinformation surrounding constitutional conventions.19

While Kittridge’s assertion that a cabinet manual cannot effect a change to a constitutional convention is true in law, my concern is that given the lack of clarity surrounding constitutional conventions in Canada a ‘ruthless’ prime minister will try, and may succeed in altering the behaviour of the constitutional actors we assume are being guided by convention.

Caretaker Governments

Another of the successes of transition, which can be credited for buttressing the independence and reputation of the public service, was the decision by the New Zealand government to identify rules for the caretaker period. These are contained in the Cabinet Manual.

A caretaker government in New Zealand, if faced with an urgent major policy decision, must consult with the incoming government and will act on its advice even if the caretaker government disagrees with this decision.20 If the identity of the incoming government is not clear, the cabinet rules in New Zealand stipulate that substantive issues are either (a) deferred; (b) handled in such a way as to avoid committing any future government; or (c) resolved via consultations with other political parties so that the action has the support of the majority in parliament.

In New Zealand, prior to the first election held under proportional representation in 1996, extensive preparations were made to minimise the number of significant issues falling to a post-election caretaker administration. Thus, decisions on major policy matters and political appointments were brought forward to before the election with potentially divisive issues identified for deferral until a new government had taken office. On budgetary decisions that could not be deferred (e.g. annual funding allocations to education institutions), final decisions were only taken following discussions with the opposition parties.21

That does not mean the system is perfect. For example, New Zealand proscribes caretaker governments from undertaking new policy initiatives or changing existing policies. The implication is that the implementation of existing policies (i.e. the policies of the government prior to the election) may continue. Yet the introduction of existing policies might itself be controversial. Discussion on issues such as these is healthy for a democracy and that can only happen if the caretaker conventions are known, like they are now in the United Kingdom, Australia and New Zealand.


The undisputable evidence from New Zealand is that comparative research into questions like constitutional conventions and government formation is indispensable. Even though this was done by parliament in advance of its anticipated divided parliament, the political class showed itself to be incompetent in applying those lessons. At the public service-level it proved to be crucial for a variety of reasons including most importantly the continuation of government services in the face of political uncertainty and ensuring the public service remained neutral and was insulated from the political discussions that led, after two months, to government formation in 1996.

There are a number of specific system adaptations that Canada might wish to consider based on the New Zealand example. A shorter parliamentary term of three years means no snap elections and no prorogations. The governor general insisting that the potential PM negotiate parliamentary support before being sworn into office (and ascertaining that this support exists) has ensured support for the government by the majority in the chamber from day-one to the end of the negotiated agreement, and this should be intuitive given that this support is the sine qua non of responsible parliamentary government.

The choice of governor general is also a lesson worth learning. By choosing a jurist to be governor general; and by his decision to educate the public about the constitutional conventions, New Zealand was able to transition through what was a politically tumultuous period and come out the other side with full confidence in the validity and effectiveness of their constitution.

During the period of government formation, in New Zealand there are no journalists camped out at Government House speculating about what the governor general might do, instead they (and the public) focus on parliament where it is understood the politicians must find a solution among themselves. Contrast this to Canada in 2008 where the possibility of the formation of a coalition government and the PM’s request for prorogation to scuttle it led to a media circus at Rideau Hall. Even now, five years later, many Canadians have little to no idea what occurred in that event or why.

Around the time the Meech Lake Accord was negotiated, I suggested a clause should be addedwhereby the Chief Justice would automatically take over as Governor General when the GG’s term (as established by convention) comes to an end. This would remove from the PM the power to recommend who should hold this office, an office that periodically may be called upon to (as Eugene Forsey used to say) ‘thwart the will of a ruthless prime minister’. It would bring gravitas to the office and insulate it from the political fray, bring the necessary legal expertise for those rare occasions when the reserve powers are being called upon for use in a political dispute and would create a scheduled turnover at the helm of the high court. The idea went nowhere but may be worth revisiting.22 (Of course, there needs to be a better process for choosing Supreme Court Justices but that is separate issue.)

It would also be helpful to have a governor general follow the example of Hardie-Boys and educate Canadians about the constitutional conventions. Failing that, written decisions when the reserve powers are used in any controversial way or an enunciated apolitical decision rule, such as that which guides the Speaker of the House of Commons when he casts a deciding vote, would be improvements.23 The New Zealand experience would seem to support these ideas.

The cabinet manual is an issue already being discussed in Canada. It is about process and content and unless the government of the day is committed to both in an impartial fashion so as to ensure that the system of government is optimized in the tradition of responsible parliamentary government (which has been the case in New Zealand, Australia and the United Kingdom) then the document will not be credible and it could even be damaging for our democracy.

In contrast, publishing the ‘caretaker conventions’ is essential and even a blatantly partisan pro-‘outgoing executive’ document would begin a much needed public debate about what should be the limits on a government during an election or upon defeat in the House of Commons.


1 Bruce M. Hicks, “The Westminster Approach to Prorogation, Dissolution and Fixed Date Elections”, Canadian Parliamentary Review 35:2, 2012, pp. 20-27.

2 Bruce M. Hicks, “Lessons in Democracy from Australia: Coalitions Governments to Parliamentary Privileges”, Canadian Parliamentary Review 35:4, 2012, pp. 25-36.

3 Often they are false majorities in that under SMP a political party can win more than 50 percent of the seats in the legislature without winning 50 percent of the popular vote.

4 Jonathan Boston, Governing Under Proportional Representation: Lessons from Europe. Wellington: Institute of Policy Studies, Victoria University, 1998, pp. 5-7.

5 Jonathan Boston and Stephen Church, “Forming the First MMP Government: Theory, Practice and Prospects” in Jonathan Boston, Stephen Levine, Elizabeth McLeay and Nigel G. Roberts (eds.), From Campaign to Coalition: The 1996 MMP Election. Palmerston North: The Dunmore Press, 2000, p. 5.

6 Ibid., pp. 104-8.

7 He has actually fired by the governor general acting on the advice of the prime minister. In New Zealand it is understood that the GG will refuse that advice if the PM no longer has the confidence of the legislature.

8 Jonathan Boston and Stephen Church, ’The Impact of Proportional representation in New Zealand on the Formation, Termination and Effectiveness of Governments’, paper delivered to the 96th annual meeting of the American Political Science Association, Washington, 2000, p.5.

9 Peter Aucoin, Mark D. Jarvis and Lori Turnbull, Democratizing the Constitution: Reforming Responsible Government.Toronto: Emond Montgomery Publications, 2011.

10 Speech at the Harkness Henry Lecture entitled “Continuity and Change” made on July 31, 1997 (available at and accessed on September 15, 2013).

11 Jonathan Boston, Stephen Levine, Elizabeth McLeay and Nigel S. Roberts, New Zealand under MMP: A New Politics? Auckland: Auckland University Press, 1996, pp. 116-51.

12 Colin James, Under New Sail: MMP and Public Servants (Victoria University of Wellington: Institute of Policy Studies, 1997).

13 Marie Shroff, The Operation of Central Government under Proportional Representation Electoral Systems: Report of a Visit to Germany, the Netherlands, Sweden and Ireland. Wellington, Department of the Prime Minister and Cabinet, 1994 [unpublished]); State Services Commission, Working under Proportional Representation: A Reference for the Public Sector. Wellington: Government of New Zealand, 1995.

14 For e.g., Jonathan Boston, Governing Under Proportional Representation: Lessons from Europe. Victoria University of Wellington: Institute of Policy Studies, 1998.

15 Jonathan Boston and Elizabeth McLeay, ‘Forming the First MMP Government: Theory, Practice and Prospects’, Jonathan Boston, Stephen Levine, Elizabeth McLeay and Nigel S. Roberts (eds.), From Campaign to Coalition: The 1996 MMP Election. Palmerston North: The Dunmore Press, 1997, pp.2 27-8; and Richard Shaw, ‘Rules or Distraction? Officials and Government Formation under MMP’, Political Science.51:1, 1999, p. 42.

16 Rebecca Kitteridge (Deputy Secretary to the Cabinet), “The Cabinet Manual: Evolution with Time”, paper presented to the 8th Annual Public Law Forum, March 20-21, 2006. Available at and accessed on September 17, 2013.

17 Ibid.

18 Cabinet Office, Cabinet Manual. Wellington: Department of the Prime Minister and Cabinet, 2001.

19 Bruce M. Hicks, “Advice to the Minister of Democratic Renewal: Senate Reform, Constitutional Amendments, Fixed Election Dates and a Cabinet Manual”, Constitutional Forum 21:2, 2013, pp. 23:37 2013.

20 Matthew Palmer, ‘Collective Cabinet Decision Making in New Zealand’, in Michael Laver and Kenneth A. Shepsle (ed.), Cabinet Ministers and Parliamentary Government. Cambridge: Cambridge University Press, 1994, pp. 244-5.

21 Jonathan Boston, Stephen Levine, Elizabeth McLeay, Nigel S. Roberts and Hannah Schmidt, ‘The Impact of Electoral Reform on the Public Service: The New Zealand Case’, Australian Journal of Public Administration 57:3, 1998, pp. 68-9.

22 While I proposed it as a constitutional amendment, this change does not require a constitutional amendment, just one PM doing it and then another one or two following suit and it will become a constitutional convention.

23 Bruce M. Hicks, “Guiding the Governor General’s Prerogatives: Constitutional Convention Versus an Apolitical Decision Rule”, Constitutional Forum 18:2, 2009, pp. 55-67.