Then years of MMP

Colin James’s notes for comments to Australasian Study of Parliament Group seminar, 5 December 2006

1. The precursor: the decline of the Nat/Lab major party vote

1951…. 99.8% (the third election it had been above 99%)

1972…. 89.9%

1981…. 77.8%

1987…. 92.0%

1993…. 69.7% after betrayals by both major parties

–Labour junked the guaranteed job

–National went radical [Second big deviation — first was Muldoon’s into populism]

The vote for MMP in 1993 was a vote against betrayal. The big-two-party vote dropped to 62% in 1996.

The converse of the fall in two-party vote was rise in number of MPs from other parties:

1966-69…. 1 [Social Credit � Hobson]

1978………. 1 [Social Credit � Rangitikei by-election]

1978-84…. 2 [Social Credit � also, two Labour MPs became in effect independents 1983-4, likewise 2 National MPs 1984]

1984-87…. 2 [Social Credit]

1989-90… . 1 [Anderton, New Labour, no by-election]

1990-92…. 1 [Anderton, first renegade from a major party to win re-election in a general election for many decades]

1993………. 2 [Peters in by-election joins Anderton]

1993-96…. 4 initially [Anderton, Lee, Peters, Henare] building to 16, via Dunne 1994, 5 United (including Dunne) 1996 and desertions from Labour and National to NZ First (2), Christian Democrats (1), Conservatives (1) and independent (1)

In summary: There was a real likelihood that even under FPP there could have been a minority government in 1996 — as did happen 1912-25, 1928-31 and as nearly happened in 1981, 1993. Nevertheless, MMP passed only by 54%-46%. [Support for MMP in Heylen poll dropped from 75% in February to 54% just before the vote.]


2. Parties since 1996

Election…. No. of small parties…. % vote…. % wasted vote

1996………… 4->5…………………… 38%…….. 8%

1999………… 5->6…………………… 31%…….. 6%

2002………… 5->6…………………… 38%…….. 5%

2005………… 6……………………….. 20%…….. 1%

This doesn’t look yet like the emergence of either the German model of two countervailing large parties with one or two smaller parties or the Scandinavian model [as some Labour people hoped for after the 2002 election cut National to 21% — but actually the real National vote was nearer the 31% it got in electorate seats].


3. MMP’s four confusions

a. The major parties took time to get the “party-vote-that-counts” message. Labour was quicker on the uptake than National which was still befuddled in 2002, with result that it had to go overboard on “party-vote-that-counts” in 2005, at some cost of goodwill among smaller parties.

b. Big parties also took time to figure that the best strategy for maximising their vote was, counter-intuitively, to concentrate in their safest areas instead of treating the whole country as a marginal seat. Labour understood this in 2005 with its strategy to get out non-voters in state house and poorer areas (though only in about half the relevant electorates). The progress of the election count on the night showed that. Under FPP concentrating on safe seats would have been stupid.

c. The new system presented tactical challenges to which the major parties offered inconsistent responses which at times (understandably) attended to short-term interests at the expense of long-term interests.

(i) Labour in 1999 failed to realise that its interests lay in knocking out Winston Peters in Tauranga, which would have taken out New Zealand First.

(ii) In 2005 National went hell-for-leather to oust Peters in Tauranga only to find him in alliance with Labour post-election and gunning for National with vigour. National was also in two minds about Rodney Hide in Epsom in 2005, some arguing that ACT took votes which would almost all be National’s if ACT was knocked out but others arguing that in the 2005 election National needed all the votes it could get, whether directly or indirectly, and ACT’s votes would be denied to the National side if ACT was not in Parliament.

Contrast Labour’s heavy hints to Labour voters to elect Jeanette Fitzsimons in Coromandel in 1999 (and, unsuccessfully, to elect National’s Richard Worth in Epsom to knock out ACT).

d. Smaller parties at times added to the confusion: ACT had the logical but impractical idea in the early 2000s that National should concentrate on electorate seats and donate its party vote to ACT; New Zealand First and United Future have generally overstated their potential for influence (though see (e)(iii)); the Greens at times in 2002 appeared to be in opposition to Labour, its logical major-party ally.

e. Voters also struggled to work out how to work the new system. The political culture in NZ had been and is to elect or un-elect a government — or to cast a dissident vote. MMP made this more complicated, especially in 1996 and 2002. Proportional representation implies (in theory) that voters elect a parliament which then sorts out a government.

Labour grasped the implication of the political culture more quickly than National (perhaps because of its galling experience in 1996 of having to fend off a challenge for primacy on the left from the Alliance). In 1998 Labour formed a coalition-in-waiting with the Alliance. Critics ignorant of the political culture, including National’s strategists, thought that would sink Labour by association. National still hadn’t quite got the message by 2005, failing to present a clear alternative government.

(i) Voters took time to get the party-vote message — some still haven’t; a notable group which did understand the message in 2005 was the large numbers of those who gave the Maori party their electorate votes gave their party votes to Labour.

(ii) Voters have struggled to balance the need for their favoured large party to get enough votes to take the initiative in forming a government with the need to get potential partner-parties over the 5% hump; in crunch electorates, however, they by and large have seemed able to work the electorate seat imperative out effectively (Ohariu-Belmont from 1996 on, Wellington Central 1996, Coromandel 1999, Epsom 2005).

(iii) Voters in 2002 worked the system effectively in considerable numbers to deny Labour a clean run by diverting party votes from National to New Zealand First and United Future when it became clear National could not form a government, only returning their votes to National when in 2005 there was a real contest.

Nevertheless, despite the proliferation of small parties, the evidence is that voters are bit by bit honing the system. The total vote for small parties has been cut from 38% in 1996 to 20% in 2005 and the “wasted vote” from 8% to 1.3%. My guess is that 2008 will see the first cut in the number of parties in Parliament, though maybe by only one party.


4. An aberration in the system is the rule giving party vote proportionality just for getting an electorate seat, thus allowing some parties to get around the 5% threshold. In 1996 the Christian Coalition got 4.3% and no seats; in 1999 NZ First got 4.3% and five seats. The logic is not immediately obvious.

In April 2002 at a conference Sir John Wallace, chair of the Royal Commission on the Electoral System and then of the Electoral Commission, said he would not recommend that if the Royal Commission was doing its work in 2002. He said: “The New Zealand voting public is unhappy and cynical about political conduct and anything which can have an aura of clever practice is better avoided. I would therefore abolish the waiver.”

Note that in 2005 we got an overhang but that it was due to the ethnic peculiarity of separate electorate seats for Maori. If the Maori party maintains its current plausibility, there might be a bigger overhang after 2008 election.

Longer term, are the Maori seats an aberration in a system which is designed to give minorities better representation and in fact has done so, especially for Maori? In pure electoral logic, they are an aberration. But that leaves aside the tangata whenua argument for retention of the seats


6. Constitutional innovations

Already we have had a number of governing variations:

–majority coalition;

–minority coalition with committed majority support on confidence and supply from another party or other parties;

–minority single-party government with majority support from another party or other parties on confidence and supply and ministers from other parties bound by collective responsibility;

–minority coalition government with majority support from another party or other parties, ministers from those support parties who are not bound by collective responsibility except in respect of their particularly portfolios and “spokespeople” from a party that is not a majority support party.

We have not yet had a minority government without committed majority support from other parties. This conceivably could happen after 2008.

Government continues pretty much as before. The underlying Westminster principles have proved flexible enough to accommodate the innovations, including officials briefing non-government parties and even opposition parties when their support is needed on specific bills.

One notable innovation after 1999 was to allow the minority partner in the ruling coalition to vote against legislation. The Alliance did this from 1999-02. The scope for radical agendas is diminished. But that was going to happen anyway after the big stir 1984-92� Reform is still possible, cf the Supreme Court Act, the Prostitution Reform Act and the Civil Union Act.


7. In Parliament:

The Parliament is more representative of women and ethnic minorities. Maori now have rough proportionality. There are Asian MPs and more Pacific MPs. This was evolving under FPP but MMP promoted it.

Note that MMP was played as “More Maori in Parliament” on Maori FM radio stations in 1993. And so it has proved — but mainly because of retention of separate Maori electorates. The 1986 Royal Commission had recommended no threshold for an ethnic party. [This was unworkable, which reflects a lack of a politically-practical person on Royal Commission.]

Do Maori have more leverage? Not obviously yet. In the crunch Maori rights claims still depend on persuasion on grounds other than numbers in Parliament (cf Foreshore and Seabed Act). Will the Maori party change that? Possibly, if it holds the balance of power after 2008. In that event, that leverage will be for specific objectives and will need to be exercised with care to avoid being blamed for instability and jeopardising the Maori electorates.

The real test for Maori parliamentary representation will come through the party vote, not the electorate vote. The logic of MMP was always for a separate grouping to hold the Maori seats. Longer-term, however, Maori need to capture the Maori party vote. The 2008 election will be an interesting first test of whether in time the Maori party can make that transition.

Small parties have yet to work out how to manage coalition partnership. There is folklore that support parties lose votes in the next election because of complicity with the government and/or failure to get the government to modify its programme as its voters would wish. But other factors have been in play as well, for example, conservative New Zealand First voters in 1996 would have preferred coalition with Labour and/or disliked the brashness of the Maori “tight five” and many United Future voters in 2002 were essentially National voters voting for restraint on Labour but not necessarily United Future’s platform. So it is not clear that being a coalition or support party alone is the key factor.

The tensions of being a coalition party or support party have caused internal damage: New Zealand First split in 1998, the Alliance split in 2002 and some United Future MPs became tow-ey in 2004-05.

Select committees are now more relevant. They do change legislation according to submissions and at times against government wishes, including the occasional junking or gutting of bills (though some of that has been game-playing).

Legislation is slower, in part of because of the need to build working majorities for each bill, which takes time and in part because the scope for urgency is limited and the business of the House is governed by the business committee, not by government whim or needs.

In some respects legislation is better because mistakes or unworkable/counterproductive clauses are picked up (especially in business legislation). It has also in this term allowed some technical bills through that have languished in the “maybe sometime” box.

Opposition parties can score success if they produce well-drafted legislation, with specific objectives.

But major parties prevail; it is their programmes which by and large are implemented. It is still valid to talk of a “Labour government” or a “National government”, even if that is not strictly correct.


8. For lobbyists

It is not enough to lobby ministers and officials only, though these remain the prime focus. It is useful for lobbyists to keep contact with all parties, including opposition parties.


9. The future for MMP

Taking a 10-year view:

Unless there is a crisis, MMP will stay. The parties and public are getting the hang of it and, except in the National party, there is no drive for change — and certainly not for reversion to FPP. Helen Clark was the lead-MP in the campaign against MMP. She has accommodated to it.

But it needs fine-tuning.

Will the threshold waiver for winning electorate seats go? Not while any government depends on a party which wins an electorate seat.

At some point the Maori electorates will go. This will be accelerated if Maori become disproportionately over-represented. [Set aside the Treaty argument for equal representation for Maori and non-Maori.] That may depend on whether the growth in the number of seats stalls, as latest census figures indicate. The logical change if the system is changed is to supplementary member, with proportionality only of the list. Labour would have had single-party governments in 1999 and 2002 if the vote had been the same under SM (but of course it would not have been).

Could there be a related change? For example, to fixed terms? Helen Clark did indicate in the runup to the 2005 election (mindful of the almost universal criticism in 2002 when she called the election early on a pretext), she might agree with fixed terms if they were four years long. The electorate has shown no taste for that in referendums in 1967 and 1990 and is most unlikely to have developed any more taste since.