Whatever it takes: Government formation after the first four MMP elections

Colin James’s Australasian Study of Parliament Group, 30 November 2005

The touchstone is a majority in Parliament. The practice has been to obtain a committed majority. And the experience has been that what starts out as the governing support arrangement in a parliamentary term is not what ends up. Government re-formation is therefore a relevant side-issue.

1. What happened

a. In 1996 there was an auction, Labour and National bidding for New Zealand First’s, or more accurately, Winston Peters’, 17 seats. National won, in part by examining what it could agree with in New Zealand First’s manifesto and offering that — plus the post of Treasurer. Labour worked the other way round, offering New Zealand First what it thought it could offer — which excluded the post of Treasurer. Labour could not have offered that and also maintained credibility with the international money markets.

National had an additional advantage: it had more seats than Labour (44 to 37) and, together with New Zealand First, it could form a bare majority, 61 seats. Labour had to obtain an endorsement of sorts from the Alliance, so with New Zealand First could make only a plurality. Those were plausible reasons for New Zealand First to choose National.

In fact, if Michael Laws, a close adviser of Peters, is to be believed (and I think he can be) Peters intended to go with National from the start. [See endnote.]

The coalition lasted only 20 months. National had been weakened in the 1993 election (to 35%) and further weakened (to 34%) in the 1996 election. There was some unease among MPs at the association with Peters, first, because he was seen as an unreliable populist renegade (having been fired as a minister in the National cabinet in October 1991 and then having left the party to form his own party) and, second, because alliance with Peters watered down the National “brand” from “blue” to “grey”. This resentment fuelled a coup against Jim Bolger as Prime Minister and his replacement with Jenny Shipley in December 1997. Shipley broke the coalition up in August 1998 when Peters tried to block the sale of shares in Wellington airport. Shipley governed thereafter with a makeshift majority of remnants of New Zealand First, which split over the coalition breakup, with support from ACT and the United (both outside the government) and a deserter from the Alliance, Alamein Kopu. Shipley saw out the full term.


b. In 1999 there was a government in waiting. Labour and the Alliance agreed in August 1998 that they would go into coalition. This gave voters a clearly visible governing combination to vote for, which they did. Consequently, coalition negotiations were completed very quickly after the election.

In fact, they were completed a little too quickly. What had been a majority on election night — Labour 52 plus Alliance 11, making 63 — turned into a minority (59 seats — Labour 49, Alliance 10) on the final count when the Greens picked up both the Coro-mandel electorate and enough party votes to clear 5% and enter Parliament after having fallen short in the election night count. In a deal which was never formally signed be-cause of animosity between the Alliance and the Greens — the Greens having left the Alliance to campaign separately — the Greens provided support on confidence and sup-ply votes, in return for some small fiscal concessions.

In formal terms the Labour-Alliance coalition lasted until Clark called a snap election for 27 July 2002. But actually the Alliance split in April 2002 when leader Jim Anderton called the bluff of dissidents in the caucus and the wider party (who were a majority) by forming his own party, Jim Anderton’s Progressive party. A fiction was invented and maintained of an “Alliance in Parliament” which, for purposes of the Electoral Act, was still united. Dissident Alliance ministers even remained in the cabinet and the ministry.

This inventiveness foreshadowed the even more inventive arrangements in 2005. It could be said, a precedent for the bizarre was set then.


c. In 2002 there was another fast deal. Anderton’s party (consisting of himself and one other MP) was quickly in coalition with Labour and a deal struck with United Future, an unlikely combination of liberal centrists and evangelical christians, for a commitment on confidence and supply votes. The Greens, who had campaigned against Labour on genetic modification, were left out of calculations, even though they also could have fur-nished a majority.

Instead Helen Clark reached right, thus potentially, so Labour acolytes hoped, establishing a “Scandinavian model” of a powerful left-based government dominating a balkanised right. Actually, this provided the cover for a “left” government: with the Greens’ support, Labour (and Anderton) passed a raft of leftish, union-friendly and social-liberal legislation. Reaching right, Clark had actually turned left. By 2002 inventiveness was becoming Clark’s middle name.

This arrangement held throughout the 2002-05 term. But United Future’s leader, Peter Dunne (a former Labour cabinet minister in 1990 who had also served in a National cabinet in 1996, to add to the confusion), leaned right during the 2005 campaign, pointedly having coffee with National leader Don Brash but not with Clark. Clark’s magic in 2002 turned out not to be what it seemed.

And actually, the governing arrangement didn’t quite hold 100 per cent. Tariana Turia left Labour to form the Maori party in protest at Clark’s Foreshore and Seabed Act. It left intact Clark’s confidence-and-supply majority (with United Future) but it foreshad-owed a hollowing of Labour’s Maori support in the electorate.


d. In 2005 the government took a whole month to assemble. Clark turned right again and became even more inventive.

In theory, she could have formed a “left” government: the Greens and the Maori party added up to 61 seats, a majority in a Parliament expanded to 121 by one Maori party overhang seat. The Greens had gone out on a very long limb to pledge themselves to a Labour-Green government and moderate their demands accordingly. But the Maori party owed its very existence to reaction against Labour’s foreshore and seabed solution and on some matters some of its prominent people were decidedly conservative. Those who imagined a “left” government were even more imaginative than Clark.

Clark might just possibly have turned left if the Maori party had been prepared to com-mit to abstaining on confidence and supply. That I have on excellent authority. But in fact the Maori party could never have done that. As a movement, rather than a party, it had to stay outside the numbers game until it could establish a position and a corporate spirit. Where it went wrong was to play footsy with National and join ACT and United Future in a bid to get New Zealand First to side with National and thwart Labour. There was just the whiff of auction in the final few days of negotiations and that had a pro-found effect on Labour’s conduct of those negotiations in the final days.

And in any case Clark had two other more powerful considerations. One was to get a viable committed majority — and the Maori party could not in any real sense have delivered that. Clark is a belt-and-braces sort of Prime Minister and a plurality of the sort Shipley had after August 1998 will not do. The other was her determination not to let the centre slip away, a determination intensified by the loss of seats and votes in the provinces in the election. So she set out to lock in New Zealand First and United Future, agreeing to a remarkable proportion of New Zealand First’s demands and nearly everything United Future asked for, even when it went back for more on the penultimate day, expecting to be rebuffed.

The difference with 2002 is that this time she has turned right and cannot run a left government. Will it last? Will there be a government re-formation during the term? The fiction that there can be ministers who are also in the opposition beggars belief. If she pulls it off, she will command a chapter in constitutional texts and histories.


e. 2008

This is by definition speculative. But if the Maori party develops into a tightly knit party and tightens its hold on Maori electorate voters, it could well hold the balance of power after the next election (when there will almost assuredly be eight Maori seats), as it would have after 2005 if New Zealand First and United Future had gone with National.


2. Numbers of parties involved

Curiously, the number of parties involved in post-election agreements has been rising: two in 1996, three in 1999, four in 2002 and five in 2005. At some point, this must reverse.


3. Who does the negotiating?

New Zealand does not have a formal go-between, as Holland has. Negotiations have typically been conducted by a very small number of people in each party: the leaders, maybe the deputy leader and, ad hoc, an MP or two, the chief of staff and, in 1996, Peters’ lawyer brother Wayne. All negotiations, at least those that have led to formation of a government, have been bilateral. This gives the leader of the large party (Labour or National) significant initiative.

Typically, the wider parties have not been brought into the picture until it is time to ratify or not. Labour MPs in 2005 outside the negotiating team were not consulted before the deals were announced. The Greens in 1999, 2002 and 2005 declared that any deal would have to be approved by a delegate conference. The Maori party held hui for guidance on whether to negotiate and over what. The New Zealand First caucus approved the deal in 2005 before it was announced.

The 1999 deal was done between Helen Clark and Jim Anderton but the broad principle of agreeing to agree had been approved by the two parties beforehand. The 2002 deal was done between Helen Clark and Peter Dunne within days of the election and before the final result had been declared, though it was not announced until afterwards.

In 2005 Labour’s team was Prime Minister Helen Clark, Deputy Prime Minister Michael Cullen and Clark’s chief of staff, Heather Simpson. Simpson did much of the detailed negotiating with counterpart parliamentary party chiefs of staff but the substantive agreements were decided by the party leaders. Clark overruled Simpson to make the offers/agreements more agreeable to New Zealand First and United Future.


4. Types of agreements

The 1996 coalition agreement between National and New Zealand First was voluminous and very detailed. It read like a remit report from a party provincial or regional confer-ence, with odd grammar and terminology.

The 1999, 2002 and 2005 coalition agreements between Labour and the Alliance and Labour and Jim Anderton’s Progressive party were perfunctory framework agreements. The important innovation in the 1999 agreement was the agree-to-disagree provision, which allowed the Alliance, for example, to vote against free trade with Singapore.

The 1999 and 2002 agreements between Labour and the Greens were also relatively simple. The 2002 agreement committed Labour to consult on land transport and, on a lesser plane, on a limited range of other matters and Green ideas were included in the government’s eventual land transport strategy and legislation. The 2005 agreement sets up three “levels” of consultation and policy commitment: energy saving and buy-New Zealand (on which the Greens are to supply “government spokespeople” and certain policy commitments are agreed); policies and legislation agreed at quarterly meetings (on environmental education, nutrition and the community and voluntary sector; and consultation on the broad direction of policy and legislation.

The big innovation is in the appointment of Winston Peters and Peter Dunne as ministers outside the cabinet, committed to support the government only on matters within their portfolios and otherwise free to oppose the government except on confidence and supply votes — a sort of “disagree to agree” provision. In return each has a raft of policy commitments from the government, ranging from action to “review”.

How well these arrangements will last is anybody’s guess. Dunne will want to be aligned with National in the next election campaign, which will require some distancing from the government before the campaign begins. New Zealand First shows signs of developing a degree of distinction between the party and caucus and Peters. This may prove difficult to manage, though the election of Dail Jones as president gives some basis to expect it will be managed competently. Much will depend on whether Peters wants to lead the party into the next election or accepts some other offer.

The extent of both the New Zealand First and United Future lists of agreed items reflects an element of auction. National got ACT and United Future’s endorsement for an approach to New Zealand First to set up a governing arrangement which would also have depended on the Maori party. This, by some accounts, unnerved Helen Clark and encouraged her to agree more readily to demands by both New Zealand First and United Future; by other accounts, even if unnerved, the National manoeuvre made her more determined to lock in both parties. In that sense, Winston Peters might be said to have reversed his 1996 tactic and used National to ratchet up his deal with Clark.


5. What part did the voters play?

In 1996 Winston Peters led most of his supporters to believe he would use their support to oust Jim Bolger and National. He made many attacks on Bolger and National and de-clared Bolger “unfit to govern”. Polling evidence suggests most of those voting for New Zealand First expected him not to support National. In 1999 the voters delivered a harsh, nearly fatal verdict: Peters and the party survived by only 63 votes in his Tauranga electorate in the 1999 election. That weighed heavily on the party approaching the 2002 and 2005 elections and led to a decision to support the party with the most votes post-election, a decision Peters adjusted in two ways to give himself more flexibility — to a commitment to abstain from supporting either and/or to support the larger party in the first instance only. This left his position unclear — except in his commitment not to accept “the baubles of office”, that is a ministerial post. When he over-rode that last commitment there was some disquiet in the party and the president, Doug Woolerton, resigned.

In 1999, 2002 and 2005 voters for the Alliance, Progressive and Greens could have been in little doubt, if they had taken any notice of election propaganda, that they were endorsing parties which would support Labour and only Labour post-election.

A high proportion of those who voted for the Maori party electorate candidates in 2005 voted Labour on the party vote, thereby sending a clear message to their MPs.

United Future voters in 2002 probably came mostly from the National side of the fence and may well have been searching for a way to constrain what they felt was an inevitable Labour government. It is doubtful they would wholeheartedly have approved of Dunne’s confidence and supply deal, as their desertion in 2005 may have underlined — though all small parties were squeezed and in any case National was a serious contestant for power and therefore worth voting for, by contrast with 2002.

Labour voters, when polled in 2005, indicated a preference for a deal with the Greens, not New Zealand First and United Future.

Essentially, the voters are passive players in government negotiations. They deal the cards for the game but have no say at all in how the players play their cards.

There is also a curious inversion of this: Clark in 2005 said, as others had before her, that she had to play the cards the voters had dealt, implying impotence at the hands of the electorate.


6. What part did Parliament play?

In essence, no part at all. Parliament has not been the arbiter, as in 1912 and 1929. It has been a rubber stamp.

Clark in 2005 might have jibbed at the demands of the minor parties and defied them to force her out. She wanted a commitment instead.

In September 1998 after the breakup of the National-New Zealand First coalition Shipley did put down a confidence motion to prove her majority. This is the only case where Parliament has had more of a role to play than a rubber stamp, though even then she had assembled the numbers in advance, even if not in all cases by way of formal agreement.


7. What part did the Governor-General play?

None. National deputy leader Gerry Brownlee twice suggested the Governor-General should inquire into the constitutionality of the Peters/Dunne ministerial appointments (perhaps taking his cue from a similarly curious suggestion by his colleague, Richard Worth, in 2003 that the Governor-General should refuse to assent to the Supreme Court Bill without a referendum).


8. What part did foreign precedent play?

In an election on the same weekend the German voters denied a majority to either the Social Democrats’ side or the Christian Democrats’ side. The response eventually was a grand coalition of the two big parties. Some faint voices were raised in New Zealand that that example might be followed but Helen Clark has repeatedly ruled that out: the only full-blown grand coalitions in New Zealand have been during the first world war and in the 1930s depression and Labour believes it would bleed too many votes to its left if it went into coalition (in fact Labour, which was to the left of the Liberals in 1915-18 and United in 1931-5, gained during those coalitions).

Winston Peters initially tried to justify his position with reference to Hans-Dietrich Genscher, Foreign Minister in successive German administrations. But Genscher was in the cabinet and his party was in coalition.

Genscher in fact swung between the two big parties in 1982. Is that a harbinger of something to come here? The last transfer of power between major parties outside an election was in 1912.


9. Conclusion

Who decides our government now? In the ultimate the voters do. But the weapon is very blunt and comes down in effect to a choice of which large party has the best hope of forming a government. Also in the ultimate Parliament does but so far it has been a rubber stamp.

Leaders decide, with small coteries — in some cases tiny coteries — of high party officials and advisers. Maybe that is what voters voted for in endorsing MMP in 1993. And maybe not.

What happens in the future will depend on (a) whether MMP is modified, (b) whether the number of parties declines and (c) how closely fought elections are — and, over-arching those factors, whether this country can recover the degree of consensus and harmony that marked the high point of one-party government from 1946-90.


Endnote: Laws, Michael, The Demon Profession (HarperCollins, Auckland, 1998), pp368ff, esp p377: “Each conversation with Peters convinced me that the partner decision had al-ready been made at least in Winston’s mind. There would be negotiations with Labour all right, but they would be for a distinct purpose to act as the ratchet that forced further National concessions.”