On the eve of the New Zealand election, two more polls signal that the nine-year-old Labour government of Helen Clark is in dire straits. The New Zealand Herald reports:
A One News Colmar Brunton Poll and TV3’s TNS Poll both have National with a respective 13 point lead and 12 point lead over Labour – two days from the election. National has more than 45 percent support in both polls, and would have the numbers, with support from ACT and United Future, to form a government. Both polls have Labour at 35 percent and below. With New Zealand First failing to make the five percent MMP threshold, the figures show Labour unable to muster the numbers to take the Treasury benches. The One News Colmar Brunton Poll showed National would have the numbers to govern even without the Maori Party – who are still being labelled as the kingmakers in this election. However Labour would lack the numbers necessary, even if it formed a coalition with the Greens, the Progressives and the Maori Party. The TV3 TNS Poll saw National up 1 per cent with 46 per cent support and Labour down 4.3 per cent on 33.1 per cent. John Key was the preferred Prime Minister on 36.4 per cent, with Helen Clark on 34.2 per cent. Both polls showed continued support for the Green Party on 9 per cent.
On the subject of New Zealand and its thresholds, I seek leave of the House to make a personal explanation. Writing in Crikey last week, Malcolm MacKerras accused me of conspiring to poison the minds of its impressionable readership with the heresy of MMP:
I fear that, following the New Zealand general election on Saturday week, I am likely to read either Charles Richardson or William Bowe singing the praises of New Zealand’s electoral system, known as “Mixed Member Proportional” or MMP. If I have read the two above-named men correctly it seems to me they belong to the view (regrettably all too common in psephological circles) that “any old” system of proportional representation is better than a non-proportional system.
I can only respond that MacKerras has indeed not read me correctly. Contrary to the prevailing modern view on this subject, I do not support proportional representation for lower houses in parliamentary systems, and believe Australia has struck the right balance in saving it for upper houses. My problem with lower house PR is that government formation becomes detached from the election itself: voters merely measure out bargaining power for the post-election horse trading that decides who actually wins, rather than casting a verdict on the government.
As for my alleged singing praise for MMP, I have two recollections of commenting on it in the past: once when I called it clunky, and once when I called it silly. As well as sharing the concerns MacKerras raised in his Crikey piece, I am also bemused by a system that entitles parties crossing a 5 per cent threshold to an instant six seats. The Green Party is polling strongly enough this time that it can probably rest easy, but in 2005 its vote was hovering around the do-or-die 5 per cent. It would surely be in everybody’s interests if so substantial a party could plan its parliamentary future without the fear that a slight shift in electoral fortune might condemn it to oblivion.
UPDATE: Two more polls. Fairfax-Nielsen points to a wipeout:
Today’s Fairfax-Nielsen poll shows National has opened up an 18-point lead – 49 per cent support compared with Labour’s 31 per cent – and a last-minute surge in support for ACT would put John Key in a position to pick his new Cabinet.
Better news for Labour from Roy Morgan:
On the eve of the 2008 Election the New Zealand Roy Morgan Poll shows National Party support at 42% (down 1%), a 7.5% lead over the Labour Party at 34.5% (up 2.5%) … Support for the Greens 10% (down 1.5%) is near its record high of 11.5%, while support for NZ First is 4.5% (unchanged), ACT NZ 4% (up 0.5%), the Maori Party 2.5% (unchanged), United Future 1% (up 0.5%) and Others and Independents 1.5% (down 0.5%).
Review by Antony Green here.
95 Comments
William, there are four polls out today. One News Colmar Brunton had National 47, Labour 35, Green 9. TV3 had National 46, Labour 33, Greens 9. Fairfax Media produced National 49, Labor 31 with ACT also increasing.
On those, Fairfax Media media produces a clear National win, National just gets a majority with ACT and United Future on the One News poll, TV3 is too close to call and Maori Party playing an important roll. NZHerald is worse for Labor because of the low Green vote.
I’m over in Auckland and have done a piece on the lead-up to the election at http://www.abc.net.au/news/stories/2008/11/06/2412194.htm.
Polls and campaigning are banned on polling day, which is why all the polls were out today.
I’m gonna write a lot about this tomorrow, but since you mentioned the post-election government formation, it’s interesting that, apart from the Maori Party, all parties this election have divided into two camps: National, ACT and UF on one side and Labour, Progressive, Greens and NZF on the other (although it’s more because National won’t have anything to do with NZF, and it’s unlikely they’ll return).
This appears to be moving towards the modern Italian process whereby most political parties fall into larger blocs. This means that the only two scenarios are a National+Allies majority or a Maori Party balance of power. What it means is that parties like ACT, UF and Greens have been honest about which party they support, but a vote for ACT isn’t simply the same as a vote for National, or a vote for the Greens the same as a vote for Labour. Rather, they are a vote to push the Nationals or Labour in a particular direction. It’s interesting that both UF and ACT are supporting National, but are running with very different agenda, and National would be very different, depending on how much it relies on UF or ACT.
So they are moving towards a party system where you have a two-horse race for government but still gives the enhanced democracy that comes from minority governments and minor parties in Parliament.
Key has said that he won’t give Peter Dunne a cabinet position in any Government he forms.
I remember reading an article a while ago where the Maori Party leader conceded it was extremely unlikely they’d support a National government.
There are two Maori Party leaders. I think you’re talking about Pita Sharples. He has suggested he leans towards Labour. Tariana Turia (the former Labour minister whose defection triggered the formation of the party) leans towards National. There is a clear favouring of Labour amongst Maori voters over National (even those who vote Maori Party), but Maori Party are bending over backwards to support National.
There could also be a situation where a National-ACT-Maori government would be much more stable than a Labour-Green-Maori government on the basis of numbers. For example, my election prediction is 56 National, 4 ACT, 1 UF and 54 Labour-Greens-Progressive. In that case, the six Maori MPs I predict could form a solid government with National, or try and support a much less stable Labour government.
Coalition and minority governments can happen with single member electorates too.
Canada
There are many parties in the UK parliament such ad Labour (including several Labour-Co-op MPs), Conservative, Liberal Democrat, Scottish Nationalist Party, Plaid Cymru, Ulster Unionist, Democratic Unionist, Social Democratic Labour Party, Respect, and Kidderminster Hospital and Health Concern (not to mention Sinn Fien who get elected but do not turn up).
The Maori party could sweep all seven Maori seats.
They would seem to be in the same bind that the Greens are in here. The Greens can threaten to give preferences to the Libs, or even go into government with them in Tas and the ACT, but everyone knows that Green voters overwhelmingly support Labor vs the Libs (as their preferences show), so it’s a fairly hollow threat – and if they DID do it, their voters would punish them next time. Maori have historically voted overwhelmingly (like, 90%) Labour, and even if most of then now vote for the Maori Party, they don’t like National any more than they did. So it seems to me that National+ACT+UF have got to get to 50% on their own if they are to win.
Why do some of these NZ polls persist with quoting their figures to one decimal place when the MOE is around 3%? This is false precision and is not warranted by the data. It’s as if they are trying to claim a superior level of accuracy. It’s poor research practice and they really should know better.
“The Maori party could sweep all seven Maori seats.”
If that does happen there would be a probable overhang of five (4 Maori, 1 Prog), all Labour leaning, so National would need 60 seats on their own requiring very nearly 50% of the vote in their own right to be able to form government with 3 ACT and UF parliamentarians alone.
I don’t know anything about the Maori seats or why people are generally assuming that they will win 5 of 7.
At least they get 6 seats once they reach 5%. And what do the Greens in Oz get with their almost 10%? NOTHING…
William I completely agree with your sentiments regarding the silliness of proportional voting in the lower house.
Martin B, there was a poll done for Maori TV on the Maori seats and this is being used as the basis for how many seats the party will win. At the 2005 election, half of those who voted Maori on the Electorate Vote then chose Labour on the Party Vote. The propsect of this happening again is one of Labour’s hopes for denying National a majority at this election.
I disagree, without proportional voting all you get is inflexible mainstream parties and an apathetic electorate
I think people are forgetting that there were two referendums on the NZ voting system. On the first ballot, voters were offered four alternatives, First past the post, preferential, Hare-Clark and MMP. There was almost no support for Preferential and Hare-Clark, so both were excluded and the binding referendum went forward between MMP and FPTP. It received 54% of vote in the face of a massive ‘No’ campiagn by the business community at an election where neither major party could get above 35%. 54% is a higher vote than John Howard’s 2PP at a ‘landslide’ Australian election. The success of the referendum compares well with several recent Candadian referendums on electoral reform, all of which have been defeated.
NZ has no states and abolished its upper house in 1951. MMP got support in 1993 because New Zealanders had got sick of being ambushed by both sides of politics, governments elected saying one thing and then doing something very different once they got control of government.
I know people say that MMP means government can’t do anything, but there are lots of New Zealanders with memories who quite like that. Yes the system has weaknesses, but the record of government’s from 1975 to 1993 is why MMP is in place.
I will be blogging the results from Auckland on Saturday night at http://blogs.abc.net.au/antonygreen/. I’ve also put a second piece up with a longer history of NZ and Australian politics at http://www.abc.net.au/news/stories/2008/11/07/2413152.htm
No 14
Look at Italy for a prime example of why proportional voting is nonsense.
How close are the margins?
The difference between an overhang of 3 and an overhang of 4 is pretty significant.
Hopefully someone can link all these pieces together from abc.net.au/elections. While I’m responsible for the linking within the elections site, when I file stories for general news, all the linking has to be done by sub-editors. It takes a few cross tasman phone calls to resolve.
neukauf17 – the USA has FPTP voting and they seemed pretty engaged this Presidential election.
FPTP voting should have stayed in place in Australia had it not been for Hughes.
Proportional has its place like the Upper House but preferential voting is not democratic FPTP enshrines the principle that each person gets 1 vote. Preferential voting gives some people 2 votes which is unfair, a person voting for a major party gets 1 vote whereas someone who votes Greens gets 2 one vote for the Greens and one vote for the ALP.
FPTP makes governments more stable and proportional voting for lower houses just shows you how useless these systems are look at Europe.
No 17
Look at the US for a prime example of why majoritarian voting is nonsense.
Quick question here. The Greens, NZ First etc have a 5% threshold to get; if they do, then the number of seats they get is (roughly) proportional to their vote. What’s that proportion? Say for the Greens, is it (% Greens vote) / (total vote, ie 100%), or is it (% Greens vote) / (% all parties who got over threshold)?
Example, with made-up numbers: Say the Greens get 5% (or 5.01%, to be safe), NZ First 4%, Maori 4%, ACT 2%, United Future 2%, Progressive 2%; 14% of the vote went to parties which didn’t make the threshold. Do the Greens get (5/100)*120 = 6 seats, or (5/86)*120 = 6.98 ~ 7 seats?
why is it?
No 17
Look at Scandinavian countries for prime example of why proportional voting works well.
No because ACT and UF win FPTP seats and so they then make the grade to possibly recieve extra seats by proportion.
no 23
two parties executing dictatorial power for 4 – 8 years and deeply dividing the socitey (of which half doesn’t even bother voting any more)
No 20, pretty engaged? what was the turnout?
Ah yes, the electorate seats – I forgot about them.
Does the vote going to parties who get neither 5% or an electorate affect the proportion, though?
No 21
Majoritarian voting is inherently democratic.
Generic Person, there is an entire literature on the subject of whether PR creates fractured party systems, or whether multiply divided socities require PR to reflect their complexity. The statement you made was as silly as the people who used to say “Look at Nazi Germany to see how dangerous Republics are” or “Look at Mussolini’s Italy to see how dangerous Constitutional Monarchy can be”.
If you read British literature on voting reform, you will see endless articles on what an appalling voting system Preferential voting is, nearly as bad as PR. In Australia, it is first past the post which is equally reviled. Countries like Belgium and Israel have run into serious problems in recent years with PR. But the question is whether the societies themselves are deeply divided, and how any other voting system could manage to reflect those deep divisions.
Belgian politics has two deep divisions, Labor versus Capital and Flemish versus Walloon. Trying to come up with an electorate system that reflects multiple cleavages in society is much harder than in a country like Australia, where the cleavages all tend to line up.
Ok, just checked the poll.
Four seats (Te Tai Tokerau, Te Tai Hauauru, Waiariki, T?maki Makaurau) show the Maori party with commanding leads.
One seat (Hauraki Waikato) has Labour with a strong lead.
Two seats (Ikaroa R?whiti, Te Tai Tonga) have Labour with slight leads, within the MOE.
So I presume people have assumed Maori and Labour will split these last two.
While I would hate to see a party manipulate the system, from that point of view the Labour candidates really should be running dead in these two seats, because the difference between a Maori party of 5 and a Maori party of 6 could be the difference between them holding balance of power.
no 19. a majoritarian system like the US where half don’t vote as both parties are too far away from their preferences and in the end one quarter decides who goverens is inherently democratic?
If the Nationals got close to 50% of the vote and the seats, it would be hard to justify not supporting them forming a government, especially if Labour is stuck around the mid-30%. Particularly since Labor will probably need the other two left parties, plus Maoris, plus maybe someone else like NZ First, just to fall over the line. If the public have made it clear they want a change, trying to patch together such a coalition would not be popular.
Belgium would probably simply collapse without PR, they even got 2 party systems along the cleavages…
No 32
That’s an argument for compulsory voting, not proportional representation.
The other side of the coin is that National are apparently confident of picking up ?hariu from Peter Dunne, but that may work against them as Dunne on his own might be a right-leaning overhang.
In most PR systems people vote without being forced to do so, because they have a choice. Look at voter turnouts in Scandinavia, people do vote deliberately.
#32 “and in the end one quarter decides who goverens ”
Isn’t that more a product of voluntary voting than the actual electoral system?
No 34
How would it collapse? Belgium was without a government for 8 months whilst alliances were negotiated which is simply a recipe for instability.
That could be a problem for the Maori Party.
Support a National govt and they alienate their own supporters.
Support Labour and they alienate a lot of other people and cause outrage against the Maori seats.
It’s the product of voluntary voting with both parties going for the median voter
Bird of Paradox, it’s an averaging system not a quota system. At the last election, each MP on average ended up representing about 18,500 votes. There were 29,500 votes for parties excluded in the cut up, and these votes just went out of the system.
It is a Sainte-Lague divisor system, more info here http://www.elections.org.nz/voting/mmp/sainte-lague.html
Once the thresholds have eliminated low polling parties, the total votes for each party are divided by a sequence of numbers, 1,3,5,7, …. This produces a big table, and the 120 seats are allocated to the top 120 quotients in this table. And the result is that if you divide the number of votes for each party by the number of seats allocated to the party, each MP for each party represents around the same number of votes.
The process used to be done as an iterative process. i.e. divide the votes for each party by the number of seats allocated so far to that party. It is the same process as doing it as a table. If 1,2,3, were used as Divisors, it is D’Hont, if 1,3,5,7 are used, it is Sainte-Lague. Modified Saint-Lague uses 1.4,3,5,7. The choice of divisors makes a difference as to whether larger or smaller parties are advantaged at the initial allocation stages.
No 30
Fair enough. But I would suggest that proportional representation accentuates social division rather than fostering unity. I would also posit that it promotes politicians who sell out their principles to gain power and that is not desirable in my view.
no39: I mean that in a literal sense, there would be more than 1 Belgium…
43: aha, the FPTP system in the US is fostering unity. Wasn’t aware of that…
No 41
What are you talking about? Apathy is not unique to two party systems. Voluntary voting only encourages people who are most interested in politics to participate, meaning there is disproportionate number of extremists from both sides who are involved in elections.
Really, your argument is one that supports compulsory voting, not proportional representation.
For what it’s worth, turnout in NZ has fallen since MMP was introduced.
Generic person, you can posit all you like, but there is truly a wealth of research on the subject, and it is not at all clear from the literature and research that what you say is true. What you have said is one of the standard lines of inquiry on the subject of party systems, and it is invariable proven neither true nor false.
Cheers Antony.
#41
I’d have argued the opposite; with voluntary voting you’re forced to appeal to the margins. With compulsory preferential voting both parties can be sure their base vote will go to them directly or via preferences, so can actively target the centre. Having two candidates trying to be as centrist and accomodating as possible (even if it is only pretending) is not exactly a bad thing.
William, NZ turnout has been falling since they stopped holding a local option referendum with each general election. There was always a high turnout to ensure the wowsers didn’t ban pubs within council areas.
http://en.wikipedia.org/wiki/Voter_turnout
Would be interesting to see which side this overview would support but unfortunately I can’t tell the voting system for each of these countries from the top of my head…
With that surprise result in the UK, is the Carr Hypothesis still a goer? Does it need to be modified to the safest pair of hands/nearest to centre/most boring and unlikely to do anything risky with the economy? And if so, what result in NZ would verify or refute it?
Well there was still a 5% swing away from Labour in the UK.
And a 5% swing in a by-election means what exactly?
ltep
5% away is a lot better than the 20% swing that’s been slaughtering them recently.
Is it not dead. Tomorrow will prove it wrong again, no?
Oz
I don’t know who’s the more boring and safe economically in NZ, such is my lack of knowledge of NZ politics.
Seems like this was a mega-safe Scottish Labour seat where the Tories get single figures and the main battle is with the SNP. Not sure it’s that representative of the general election.
“I don’t know who’s the more boring and safe economically in NZ, such is my lack of knowledge of NZ politics.”
Well the Nationals are considered more economically right, however some of their policies during the election seem to draw parts of policy from the left, with Labor warning of the Nationals’ “hidden agenda”.
I suspect the Nationals are a bit more left than normal on the economy due to the general drift toward the left economically throughout the western world at the moment.
So what the Conservative Party wasn’t the main opposition in that by election. If the swing is on against a government in a big way it would show up. 5% in a by election against a very unpopular government is a great result for the government and there is no getting around that.
I thought that nearly all elections in recent times show that “safe with the economy” is a threshhold; no party with adequate rating will win sufficient support for government. Once the theshhold is crossed, however, the criterion is no longer determinative; perception as “most safe with the economy” is no guarantee of success.
Trying to assess uniform swings from by-elections. What has this place come to?
Martin B
The Carr hypothesis is that the GFC will lead to changes in the way countries vote to favour centrists with a safe economic platform. The right and left will die. He put it much better than me. So it’s more than a threshold. It’s the primary concern.
Martin B 63 – Who is trying to do that?
I am doing this calculation for no particular reason but where else can I post it?
If list parliamentarians continue to be added by the formula until the last overhang electorate MP is allocated, then the current state of parties in the beehive would be:
Labour: 66 (+16)
National: 65 (+17)
NZ1st: 9 (+2)
Green: 9 (+3)
UF: 4 (+1)
Maori: 4 (+0)
ACT: 2 (+0)
Prog: 2 (+1)
Forty extra seats for almost no change in the bargaining power; and that was a close election. The chance of overhangs creating a bargaining imbalance tomorrow are still remote.
So one really can’t expect overhangs to be a problem.
No-one really
I merely agree with not reading too much into the precise size of the swing in Glenrothes particularly as to whether or not it is quantitative support for some vaguely defined theory…
Sounds implausible.
Hi William – Apologies for not reading this post before writing today’s Crikey piece on NZ (very careless of me). I don’t know why Malcolm fingered you as a fan of MMP or PR in general. I agree with your point about the sudden jump from 0 to 6 seats; that’s why I’d reduce the threshold from 5% to 3%, altho in an ideal world there’d be no threshold at all.
Martin B
Perhaps “wither” would be a better term than “die”. I haven’t done it justice. I think it’s part of this “militant centralism” movement, where the Centre are damn sure they are correct, and the Left and Right are wrong, instead of being a wishy-washy compromise of left and right policies.
Hmm, I hope that advocating tactical voting isn’t an NZ election crime
“I think it’s part of this “militant centralism” movement, where the Centre are damn sure they are correct, and the Left and Right are wrong, instead of being a wishy-washy compromise of left and right policies.”
As I was arguing the other day, these terms are largely relative anyway. The distinction between those emphasising the social aspects of the economy and hence more concerned to invest in social development for economic growth, and those who see the economy in terms of more narrowly fiscal indicators doesn’t appear likely to disappear anytime soon. If anything the GFC has just accentuated the former. OTOH I’m not expecting all merchant bankers to abandon the conservative side of the economic ground.
Has anyone mentioned this calculator on the nz election website
http://www.elections.org.nz/calculator/
interesting to play with, may try with live results
“interesting to play with, may try with live results”
Hope they’ve got a good connection
The result of the by-election would not be irrelevant to many of the contests in Scotland.
http://en.wikipedia.org/wiki/Results_of_the_United_Kingdom_general_election,_2005#Scotland
Have they ever considered incorporating preferences into the electing of electorate MPs? (whether in NZ or elsewhere)
When I voted in the London elections earlier this year, it struck me as odd that you get to state a 2nd preference for the mayor but not for the local assemblyman.
I guess it doesn’t matter so much with MMP, as what you lose in the swings you pick up on the roundabouts.
Betfair has Labour at $6.40 at present and looking at this article there is not much in this election. Very tempting for the small change.
http://www.tvnz.co.nz/view/page/484445/2263783
Scuse my going on about the Scottish result, but… that’s an absolutely dreadful result for the Lib Dems. 12.7% at the last election, 2.6% at the by-election. What happened there?
Thay voted tactically for the SNP to try to defeat Labour.
Terrible idea.
You want people who manage to get >1% of the vote nationally a seat in parliament? Every single group or organisation would incorporate themselves into a political party and manage to get a fair chunk of votes.
Look what happens in Israel. Every fringe group or extremist manages to get a seat or two and you get the balance of power being shared with people who got a tiny fraction of votes. Their threshold is 2% now but they’ve increased it a few times from 1%.
No I can’t see a reason to arbitrarily set a threshold. If people choose to wish to be represented by so-called extremists then that’s that.
No, it’s not. Because then you get your Fielding like scenario except it’s in the House where Government is formed. So someone with 0.8% of the vote ends up deciding which party governs.
Except replace Fielding with the “National Front for the Liberation of Dingos”.
Fielding is a legitimately elected senator. Why he should have any less power is beyond me.
If you look at the reasoning for proportional representation, which is that parties should be represented in the parliament in a roughly equal proportion to the vote they receive, it doesn’t make sense for there to be a threshold. It defeats the whole purpose of the system.
Otherwise you just have an disproportionate system, such as we have.
There’s different levels of disproportion-ness. There’s a formula for calculating it that I can’t remember right now but gives a PR system like NZ’s with a 5% threshold a rating of 1.1. Pretty damn close. I don’t even want to think what Australia’s would be.
If you’re saying a PR system as to achieve perfect 1.0 proportionality then you’d have to increase the number of seats to the number of significant figures of the percentage vote since there’d be an artificial threshold anyway. Eg. in Australia our threshold would be 0.7% because of the number of seats.
Yes, there’s nothing wrong with the artificial threshold that I can see. If you receive less than that you are guaranteed less than one seat in the parliament.
For those among us who doubted the importance of UK Labour’s win in the latest by-election this article puts it all into perspective.
http://www.theage.com.au/world/poll-win-a-rare-fillip-for-brown-20081107-5k7r.html
(Sorry William, not sure which thread this should go on seeing that this topic has been discussed here.)
Okay, nobody has really made any predictions as yet, so here goes:
http://www.csec.com.au/nz_pred.htm
If you couldn’t be bothered linking, I’m predicting a National minority government with ACT and United Future, with 63 seats in a 122 seat chamber comprised as follows: National 57, Labor 43, Greens 10, Maori 5, ACT 4, UFNZ 2, Progressive 1 and NZF 0 on major party primaries of 46/34.
“If the election is close, any overhang caused by the Maori Party is likely to be significant in determining the formation of government. It would be difficult for the Party to support Labor if the gap between the two major parties is too significant; however, they will be in a strong bargaining position to extract concessions from either party in their pursuits of a workable majority”
A very interesting prospect and I agree that it would be hard for the MP to support Labor if there is a wide gap in the primary (which you would have to suspect would be the case). I still suspect, however, that such support would result in a downturn in their vote at the next election.
Will William be doing live blogging of the NZ election?
Is there a live TV stream and/or radio stream anywhere that will be available online?
You’re looking for the Gallagher index. I just mocked it up for the last Australian federal election – it’s 10.33 for the House of Reps, and 6.96 for the Senate (counting only the 40 senators elected in 2007). Not great compared to NZ (1.1 for the 2005 election, in the linked article), but then it’s a heap better than Singapore. Their electoral system is flat-out scary.
3 News has live streaming from 7 tonight.
http://www.3news.co.nz/VideoBrowseAll/3NewsLiveStream/tabid/387/Default.aspx
Gotta love Singapore where you get told by the Government that if you vote for another party they will let your services go and let you live in a slum.
So will TVNZ here: http://tvnz.co.nz/view/news_minisite_index_skin/news_election_08_group
No idea which will be better coverage…
New thread (so yes, in answer to Swing Lowe’s question).