Westminster majoritarian model of democracy

The similarities between the Westminster system and majoritarian democracy are such that the terms can be used almost interchangeably (Lijphart Reference Lijphart2012). The best example of this was New Zealand prior to a change in their electoral system in 1996, with their political system once described as being ‘more Westminster than Westminster’. Lijphart (Reference Lijphart2012: 10–16) identifies five defining characteristics of the traditional Westminster majoritarian system in relation to the executive and parties, which in recent times have weakened in all four countries.

1. 1. Concentration of executive power in one-party or bare-majority cabinets. Lijphart observes how the UK political system has produced two principal parties of equivalent strength, which results in narrow one-party majorities. In the post-war period there have been only two minority Labour cabinets in the 1970s, and from 1977 a short-lived ‘Lab-Lib Pact’ existed whereby the Liberals supported the government on matters of confidence in exchange for enhanced consultation. Yet after the 2010 election, a coalition government was formed between the Conservatives and Liberal Democrats, with a formal agreement and shared cabinet posts. In the devolved parliaments of Scotland and Wales, the mixed-member proportional (MMP) representation electoral systems have resulted in several minority and coalition governments since 1999. The year 2011 was the first time that each had majority governments.

As can be seen in Table 6.1, in the post-war period most cabinets have been comprised by ministers from one party. While at first glance Australia appears to be an exception, the ‘coalition’ between the Liberals and the rural National party is so strong that for all intents and purposes they can be treated more like factions of the same party. At a national level they have been in coalition for all but two brief periods, which were while they were in opposition, the last time in 1987. They generally do not challenge each other’s sitting members at elections, and they run joint-tickets (electoral lists) for the Senate in many states. At a sub-national level, they have now merged in the state of Queensland and the Northern Territory. The federal coalition agreement rests on a brief, private letter signed by the two parliamentary party leaders, and from 1949 to 1987 was only signified by a handshake (Costar 2011). The main issues have been around allocating ministerial portfolios in proportion to seats held (and sometimes the portfolio area). The Nationals’ leader becomes the deputy prime minister. They have also generally held the portfolios of transport and regional services, trade, and agriculture. Policy issues have generally not been a feature of Australian coalition agreements, unlike in other countries.

At a sub-national level, minority governments became more common in the 1990s and 2000s, with independents and minor parties holding the balance of power in the lower houses of every jurisdiction at some point. The national parliament finally followed this trend in 2010 when Labor formed a minority government with the support of the Greens and independents. Labor signed separate agreements whereby they committed to a range of policy initiatives, parliamentary reform and privileged consultation in exchange for supply and qualified confidence. The qualification was that each Green and independent signatory was free to move or second a no-confidence motion but had to oppose any no-confidence motion moved and seconded by others.

Canada has the longest history of minority government, with eleven minority governments formed after elections since confederation and a further two formed in between elections. Yet coalition government has been more elusive (only one since confederation) despite the comparatively larger number of significant parties in the lower house. As Andrew Banfield discusses in Chapter 10, where a single party has not secured a majority in the lower house, the custom is that the party with a plurality of seats is given an opportunity to form a government. In doing so, the party must be confident that opposition parties will not pass a vote of no confidence, which is a negative rule of government formation. Yet opposition parties have rarely united in this way, so single-party minority governments have not been uncommon. These minority governments have at times relied on informal, ad hoc, shifting or loose ‘alliances’ with other parties to ensure their survival.

Minority government has been less common at the sub-national level: the province of Alberta has never had a minority or coalition government, while the provinces of British Columbia, New Brunswick, Newfoundland and Labrador, and Prince Edward Island have had single-party majority governments in recent decades. In the other provinces, there have been single-party minority governments, but rarely with any kind of formal agreement. After the 1985 Ontario election, the second-placed Liberals reached an ‘Accord’ with the third-placed New Democratic Party on confidence and supply for two years by agreeing to a series of legislative proposals. After the 1999 Saskatchewan election, the New Democratic Party formed a coalition government along conventional lines with the small Liberal Party.

As Grant Duncan discusses in Chapter 12, majority Labour or National governments with clear electoral mandates were the norm in New Zealand politics from 1935 until the introduction of a new proportionally representative electoral system in 1996. Unlike many other countries, there are no constitutional rules in relation to who manages or leads inter-party negotiations, what the process is and how a government is confirmed by the legislature (Boston Reference Boston2011). Thus, a ‘free-style bargaining’ environment has developed, and in turn, different variations and experimentations have occurred (Boston Reference Boston and Miller2001, Reference Boston, Levine and Roberts2007). After the 1996 election, a more traditional and straightforward coalition was formed between the right-of-centre National Party and the small populist/nationalist New Zealand First Party, which together held enough seats for a bare parliamentary majority. The written coalition agreement explicitly referred to the need for consensus. For example, Part B acknowledged that: ‘The experience of similar democracies to New Zealand with coalition governments is that policy is developed by consensus it being recognised that co-operation is critical to the long term stability of coalition government’.Footnote ¹ Part C stipulated that: ‘Development of policy is achieved through consensus and good sense’. Duncan details the breakdown of the coalition and the confusion over the operation of conventions in Chapter 12.

After the 1999 election, left-of-centre parties Labour and the Alliance formed a minority coalition government, relying on the Greens for confidence and supply in exchange for enhanced consultation but without a formal agreement. After the following two elections, Labour again formed minority coalition governments, this time with the small left-wing Progressive Party. In that coalition’s first term, an agreement with the centrist United Future party provided a parliamentary majority on issues of confidence and supply, in exchange for a commitment to a policy programme related to families, along with enhanced consultation. By the coalition’s second term, United Future negotiated a ‘ministerial position outside cabinet’ for its leader, in addition to a broader set of policy commitments and more consultation. Still short of a majority, a similar agreement was negotiated with New Zealand First, as both it and United Future were hostile towards Greens’ participation in government. Instead, Labour and the Greens negotiated a separate agreement whereby some policy concessions were made: two Greens became ‘Government Spokespeople’ in exchange for abstaining on confidence and supply votes to guarantee the government a working majority after the appointment of the speaker.

The most recent elections have seen National form minority governments under a similar arrangement. The ACT, United Future and Māori parties accepted ministerial positions outside cabinet and received policy commitments and close consultation by pledging confidence and supply. Interestingly, National only needed the support of one minor party after each election for a parliamentary majority but opted to include more parties. After the 2008 election they also signed a memorandum of understanding with the Green Party to identify policy areas where they could potentially cooperate. Boston (2009) first identified seven forms of multi-party governance arrangements in New Zealand following MMP, to which Boston and Bullock (Reference Boston2010) added an eighth. My own typology of governing arrangements in Australia (2013) is overlaid to present ten types of minority and multi-party executive arrangements in Westminster democracies in Table 6.2.

1. 2. Cabinet dominance. During periods of majority government, the cabinet dominates the parliament due to strong party discipline. This is no longer the case in New Zealand, and after a series of experiments with and evolutions of multi-party governance arrangements, cabinet power has been curtailed. Significantly, the National government has opted to give minor parties influence even when they do not require their votes, recognising the longer-term advantages for them in pursuing more consensus-style arrangements. This is discussed further in the next section, as conventions have come to the fore.

Maintaining dominant cabinets has long been more of a challenge in Canada, where the threats and actual instances of no-confidence votes are not uncommon. Again conventions around what constitutes no confidence have become the subject of debate. For example, there has even been the bizarre situation of the minority Liberal government losing a vote over an amendment to the Income Tax Act, before successfully moving a subsequent vote to clarify that such a defeat of a ‘money’ bill did not constitute a no-confidence vote as conventionally understood.

Cabinet dominance in Australia is partially tempered by a powerful upper house where the government usually lacks a majority. Yet public displays of party unity can mask fractious internal dynamics where cabinet is often challenged. Again, Australia provides the starkest example of this, with the most recent Labor governments changing prime ministers twice and frequent cabinet and caucus leaks with the aim of destabilising leaders and their supporters.

1. 3. Two-party system. The two-party system is in decline in all four countries, as can be seen in Figure 6.1. In Australia, until the 1987 election, the major parties received over eighty-five per cent of the primary vote, and often more than ninety-five per cent. At the 2013 election the combined vote had fallen under eighty per cent for the first time, and three minor parties held seats in the lower house, in addition to independents.

Figure 6.1: Proportion of the vote of the two major parties

In New Zealand, almost one hundred per cent of the vote was once shared between the two major parties, which were often the only parties represented in the parliament. In the first election using the new electoral system in 1996, their combined share of the vote was only sixty-two per cent. At the 2014 election their share was seventy-two per cent, but the number of represented parties had increased to seven.

The Canadian major party vote shares have also been erratic, yet the shares of over eighty per cent in the late 1940s and the 1950s have never been repeated. In the 1960s, 1970s and 1980s, the shares were around seventy-five per cent, falling to fifty-seven per cent in the 1990s and just fifty-three per cent at the 2000 election. Throughout the 2000s the shares had increased to around sixty-five per cent, and at the 2011 election, the figure was seventy per cent. The number of parliamentary parties has remained fairly static at four to five, but the present-day Conservative Party has experienced several incarnations. Its predecessor, the Progressive Conservatives, only won two seats at the 1993 election.

In the United Kingdom, the earlier elections saw the two major parties sharing about ninety to ninety-five per cent of the vote, before falling to seventy-five per cent in 1974 and below seventy per cent for the first time at the 2005 election. At the 2015 election the major parties received sixty-seven per cent of the vote, with three parties contesting government.

1. 4. Majoritarian and disproportional systems of elections. While the devolved parliaments adopted more proportional electoral systems, Westminster has retained plurality voting using single-member districts. As part of the Conservative–Liberal Democrat Coalition Agreement, there was a referendum to change to an Alternative Vote system, as used to elect the Australian lower house. The proposal was resoundingly defeated, but proposals to democratise the House of Lords using proportional representation remain on the agenda for electoral reform.

Australia’s system of preferential voting for lower house seats still gives the major parties a disproportionate share of the seats, but increasingly minor parties and independents are making it to the final count and even winning seats. Symbolically, the independent Australian Electoral Commission now refers to final count once all preferences have been distributed as the ‘Two Candidate Preferred’ rather than the traditional ‘Two Party Preferred’ vote. In the Senate, proportional representation by single transferable vote makes government majorities difficult to achieve. Aside from a brief period from 2005 to 2008, minor parties and independents have held the ‘balance of power’ in the Senate since 1981.

New Zealand has changed to a more proportional electoral system, while Canada has retained the UK model, along with similar calls for electoral reform.

1. 5. Interest group pluralism. On this criterion, all four countries have largely retained a pluralist interest group system, where interest groups compete in an uncoordinated manner. However, the traditional roles for unions, and in some respects farmer groups, have changed, while there has been a rise of postmaterialist interest groups, particularly environmental groups. Postmaterialist issues have often been at the centre of the policy platforms of third-party challengers, and many power-sharing agreements have included such issues.

The Westminster majoritarian model of democracy still exists in some form in all four countries but has had to adapt as minorities have challenged the status quo. The next section more closely examines the recent instances of minority and multi-party government to identify the most contentious conventions.

Multi-party governance arrangements

In Westminster-derived parliamentary democracies government formation has few formal rules, with no formal investiture vote required or clear protocol for who has first right to form a government, so the incumbent government occupies a privileged position, remaining in office until faced with a loss of confidence (Strøm, Budge and Laver Reference Strøm, Budge and Laver1994: 311; Paun Reference Paun2011: 447). The head of state does not initiate government formation, in terms of designating a formateur or informateur. Where the incumbent government retains their majority, they continue to govern; if they lose their majority and the opposition gains a majority, the incumbent leader resigns and advises that the opposition leader be invited to form a government. Where both the government and the opposition fail to win a majority, conventions can be tested.

The two major choices are either a coalition that binds partners to a joint policy platform negotiated during the government formation phase or a pure single-party minority government that requires ongoing negotiation, but international experience shows several variations of these arrangements (Paun and Hazell Reference Paun and Hazell2010: 215–16; Paun Reference Paun2011: 448–9). Collective cabinet responsibility is seen to favour policy-compact coalitions (Strøm, Budge and Laver Reference Strøm, Budge and Laver1994). There are risks for minor parties and independents in entering coalition-type partnerships explaining a preference for minority government formation (Curtin and Miller Reference Curtin and Miller2011: 8).

Collective cabinet responsibility has been challenged by multiparty governance arrangements, but actually has remained largely intact with the emergence of what is commonly known internationally as the ‘New Zealand’ model. In this model collective responsibility is loosened and support parties can gain ministerial roles while maintaining the right to dissent on certain issues (Paun Reference Paun2011: 450). Yet like so many presumed New Zealand innovations, the experimentation with conventions began in Australia (although New Zealand went much further).

The first stage of the ‘Australian’ model began in the early 1990s and continued into the 2000s. Unlike minority governments in Canada, the major parties negotiated with minor parties and independents and attempted to reach agreements – ideally in writing – where guarantees of confidence and supply were exchanged for commitments on political and/or policy reforms.

Independents have had written undertakings with minority Labor governments in Queensland, Victoria, South Australia, the Northern Territory, and a minority Liberal government in New South Wales. Every written agreement has the stated aims of stable government and accountability. For the government, the main issue was ensuring supply and confidence, while placing conditions under which a motion of no confidence could be supported; generally it could only be moved by the independent member (rather than the opposition) and only due to corruption, mismanagement or similar wrongdoing. In most agreements the government also pledged to run full term or to introduce fixed terms, which is a lasting consequence of many of these agreements in many states. In New South Wales, Victoria, and at the Commonwealth level, most of the independents negotiated and signed the agreements as a group. Earlier agreements with independents had focused more on parliamentary and other accountability reforms, along with promises to be consulted, committee positions, and extra staffing and resources.

The first significant deviation from the traditional Westminster practice of collective cabinet solidarity occurred in the small jurisdiction of the Australian Capital Territory (ACT) in 1998. This is unsurprising given the proportional representation electoral system, history of non-majority governments, and other non-traditional Westminster features such as the absence of a vice-regal post and necessity of an investiture vote.

Prior to the appointment of independent Michael Moore to Chief Minister Kate Carnell’s Liberal cabinet, the government commissioned Professor Philip Petit to investigate the basis upon which a non-government member could become a minister. In his report, Review of Governance of the Australian Capital Territory, Petit identified three conditions for such an appointment: Moore had to give prior notice of the issues on which he reserved the right to dissent in public and in the legislature after absenting himself from the relevant cabinet discussion (a long list of broad-ranging issues); he had to implement all cabinet decisions related to his portfolio areas, even if he had absented himself and dissented; and he could not threaten to resign in cabinet negotiations.

For Carnell, she was able to secure her position and add a more ideologically congruent voice to cabinet to balance the right wing, a tactic many have argued that UK Prime Minister David Cameron used in encouraging a coalition with the Liberal Democrats. Moore’s health portfolio was also a more challenging portfolio for the government, and while Moore gained ministerial influence, he could also be blamed for failings. Moore emphasised that the success of the arrangement was highly dependent upon trust.

South Australian Labor Premier Mike Rann adopted a similar arrangement, but granted further dispensation to the non-government ministers. Under the agreements with the independent and National party ministers, they did not have to provide a list of issues that they could disagree with, however broad-ranging, and instead were effectively given carte blanche. They were also offered a ministerial position in a future Labor government if they served in line with the agreement.

The Western Australian Nationals copied these arrangements in entering a ‘partnership’, rather than a coalition, with a minority Liberal government in 2008. Notably, the Liberals had already secured the support of two independents without the need for formal agreements, and one entered the cabinet under traditional conditions. She was subsequently demoted, but did not bring down the government. While cabinet solidarity was varied in these arrangements, cabinet secrecy was still expected.

The ‘Greens’ first entered a parliamentary accord with Labor in Tasmania in 1989 while still known as the Green Independents. It was the first such written agreement focusing on parliamentary and accountability reforms, increased consultation and resources for non-government members, but also included many policy commitments. It broke down in acrimonious circumstances and, with Labor reluctant to govern with Green support after the 1996 election, the Liberals governed without a majority or an agreement with the Greens. The major parties again refused to cooperate with the Greens during the 2010 state election, and after they won the same number of seats, Labor premier David Bartlett advised the governor to commission a minority Liberal government as they won the most votes. Governor Peter Underwood released a statement, and under the sub-heading ‘Irrelevant matters’ stated that:

Eventually, Bartlett continued as premier after inviting Greens into the cabinet. Like the Australian Capital Territory (ACT) arrangement, the Green ministers were required to provide a comprehensive list of matters of conscience where they would likely disagree, but were able to add to the list at a later date. There was also provision for ‘matters of significant concern’, which was open like the provisions in the South Australian and Western Australian arrangements, except it was framed in terms of party policy and acknowledged that Greens ministers were subject to their own internal party dynamics. The key difference from earlier arrangements was that resignation was acknowledged as an option, with the expectation that the Greens would if they could not abide by the other traditions of cabinet. While popularly described as a coalition, the other three Greens were not part of the government and could take a different position to their colleagues in cabinet. The Greens were eventually expelled from the cabinet prior to the 2014 election.

The Greens charted a similar path in the ACT. They had previously supported minority governments in the investiture votes, without entering formal agreements. After winning a record number of seats in 2008, they entered a parliamentary agreement to support a minority Labor government in exchange for parliamentary reform and one of the most extensive policy programmes contained in an Australian agreement. This was despite the Liberal leader offering cabinet positions. After the 2012 ACT election, despite losing three of their four seats, the Greens sole member entered cabinet and was explicitly subject to cabinet confidentiality but not solidarity on matters of serious concern. However, most of the agreement was, focused on specific policy commitments.

New Zealand has similarly experienced incremental changes in multiparty governance arrangements, beginning with traditional coalition and loosening in recent times. The first ‘loosened’ arrangement occurred after the 1999 election, with the coalition agreement between Labour and Alliance affirming the convention of collective cabinet responsibility.Footnote ² However, in one of the longest parts of the agreement, exception was made for ‘party distinction’:

Boston and Bullock (Reference Boston2010) contend that the previous coalition’s tight discipline proved to be flawed due to inflexibility and inability of the minor party to differentiate itself, which led to this ‘agree to disagree’ provision. However the Alliance still struggled to maintain their identity and eventually split, with its leader forming the Progressives and again entering a coalition with Labour after the 2002 election. By this stage the Cabinet Manual had been updated to reflect changing governance arrangements, and thus a shorter provision on ‘maintaining identity’ was inserted into the coalition agreement:

Labour and the Progressives then negotiated a confidence and supply agreement with United First, while Labour reached a ‘cooperation’ agreement with the Greens, in exchange for limited policy commitments. Both agreements bound United First and the Greens to publicly support the process and outcome of their own policy initiatives. After the 2005 election these agreements went further, and in doing so established a distinctive ‘New Zealand model’. There were two elements: acceptance of ministers outside of cabinet being free from collective responsibility except for matters within their portfolio areas, and the establishment of a new class of ‘minister’, known as government spokesperson.

The leaders of the United Future and New Zealand First parties became ministers outside of cabinet and were ‘bound by Cabinet Manual provisions in respect of any areas within the[ir] portfolio responsibility’ and required to ‘fully represent the government’s position’.Footnote ⁴ The ‘cooperation agreement’ with the Greens identified three levels of issues, with level one fully involving the Green Party spokesperson in the detailed development and implementation of policy proposals, thus becoming government spokespeople.Footnote ⁵

The National Opposition Leader Don Brash was predictably critical in describing the arrangements as ‘concerning from a constitutional point of view’ (cited in Boston and Bullock Reference Boston2010: 359). The party’s position changed, however, by the time they won government at the following election, after the new approach proved to be practical and politically feasible, in addition to widespread acceptance from constitutional authorities and the media (Boston and Bullock Reference Boston2010). The government was also creative in its management of potential areas of disagreement, carefully allocating the ministries and even removing contentious portfolios, while also narrowly interpreting what constituted a portfolio area (Boston and Bullock Reference Boston2010). Significantly, Boston and Bullock identify the development of new ‘informal rules’ that facilitated the new arrangements, that arguably could be considered new ‘conventions’: firstly dissent within the executive would be limited to matters of party distinction, rather than ministers being able to take independent individual positions; secondly, the aim was always consensus with dissent accommodated only as a last resort; and thirdly, any criticism of government policy would be relatively muted compared to the opposition. As previously mentioned, when National won the 2008 election they opted for these arrangements with three parties, rather than just a traditional coalition with ACT. National even pioneered a new arrangement by negotiating a ‘memorandum of understanding’ with the seemingly ideologically incongruent Greens.

Stable yet adaptable conventions

The general trends in each of these two countries provide an interesting contrast. In Australia the trend had been for minor parties and independents to incrementally gain more and more influence before finally gaining cabinet positions and then securing more and more opt-outs from collective responsibility. In New Zealand the starting point was full, traditional coalition with tightly prescribed agreement that has gradually loosened to protect cabinet’s collective responsibility while accommodating or co-opting minor parties. Yet at the core, New Zealand has preserved traditional understandings of collective responsibility, while pioneering other innovative executive arrangements under cabinet; in Australia collective responsibility has been more compromised.

The United Kingdom has maintained the more traditional approach to collective cabinet responsibility. The formal agreement between the coalition parties stated that ‘There is no constitutional difference between a Coalition Government and a single party Government, but working practices need to adapt to reflect the fact that the UK has not had a Coalition in modern times’. The deputy prime minister (the leader of the Liberal Democrats) had almost equal oversight as the prime minister. Each government department contained ministers from both parties to ensure that neither had complete control. The coalition agreement contained a comprehensive policy programme, noting likely and allowable disagreement on higher education funding and nuclear power plant construction. In this way, the coalition preserves the doctrine of collective responsibility by specifying minimal exemptions (Paun Reference Paun2011: 451). The Cabinet Manual reflected this by noting the collective decisions of cabinet can be ‘explicitly set aside’.Footnote ⁶ In the preface, Cabinet Secretary Sir Gus O’Donnell observed that:

The conventions of executive government have evolved as the influence of third parties, minor parties and independents has grown.

The Westminster parliamentary system of an executive drawn from and accountable to the legislature remains stable, and the first minister and control of the treasury remains within one of the two major parties. Concessions have been made to the challenging actors, acknowledging the democratic and popular will of the people, but real power has remained within two parties in each of the four countries, while the challenger’s influence has often been fleeting (see Brenton Reference Brenton2013). This has not stopped new challengers, nor necessarily arrested the longer-term decline in the major parties’ share of the vote: the minor changes to conventions have provided a safety-valve ensuring their fundamental stability.