Kiwiblog

Last election Labour got 2% more support on the party vote but it was a near exact tie on the electorate vote with National and Labour not just winning 31 seats each, but receiving a near identical proportion of total electorate votes with National 40.38% and Labour 40.35%.

Now I have put together in excel a little pendulum to calculate what sort of margin is needed in the electorate vote for a seat to go from Labour to National, assuming a standard swing. Of course you never have a standard swing, but it is the only thing to go with. I am using the 2005 election results but adjusted to take account of the 2008 boundaries.

How I have done the calculation is if for example National has a 6% lead in the electorate vote, I take that to be that National gains 3% (half of 6%) at the expense of Labour. You then divide 3% by the overall National electorate vote last time of 40% and apply it to the actual National vote for that electorate.

The bottom line is it means one can look at a poll result, and see what seats would change hands based on the gap between National and Labour on the electorate vote. Now only One News Colmar Brunton publish specific electorate vote results, and their last result was National 52% Labour 34% or a lead of 18%. For other polls one can take the party vote margin as a rough guide to the likely electorate vote margin, as the two tend to follow each other quite closely unless there is a freakishly low party vote result like for National in 2002.

So what are the the seats most vulnerable for Labour, to National. In order they are:

1. Taupo -2.6%
2. Rotorua -1.1%
3. Otaki 0.9%
4. Hamilton West 3.2%
5. West Coast-Tasman 4.1%
6. Palmerston North 13.2%
7. Auckland Central 14.1%
8. New Plymouth 14.8%
9. Waimakariri 16.6%
10. Port Hills 17.2%
11. Wellington Central 17.6%
12. Maungakiekie 22.2%
13. Hutt South 24.6%
14. Waitakere 25.8%
15. Mana 26.6%
16. Rimutaka 28.8%
17. Dunedin North 29.2%

The negative figures for Taupo and Rotorua indicate they are already National in terms of a paper majority on the new boundaries, and that Labour would need to score more votes nationwide than National to retain those seats on a standard swing.

A 10% electorate vote lead for National would see five seats go to National on a standard swing. At 15% lead it is eight seats and at 18% it is 11 seats. Now on top of those seats, Labour also look likely to lose two (arguably three) seats to the Maori Party so if there is a 18% gap in the electorate vote, Labour would be reduced to 17 or 18 electorate seats and National would have 43 electorate seats.

Now again, swings are never standard. Candidates and regional variations all have an influence. But the pendulum does show what seats may be first to swap hands.

As to be expected under MMP, most focus in on the party vote. But electorates are also important to parties. It hurt Labour badly to lose 10 seats to National in 2005.

Currently National and Labour have 31 seats each. And they got almost the same share of the electorate vote in 2005 – 40.38% for National and 40.35% for Labour.

Now seats never quite fall in a linear way, but one can apply a linear swing to majorities, to get some idea of what seats may fall to National if the current polls continue.

The last Colmar Brunton poll has the electorate vote for National at 48% and for Labour at 35%. This is a relative increase in National’s EV of 19% and a relative drop for Labour of 13%.

So which seats (new boundaries have been taken into account) would on paper go to National, if that was the linear swing on the electorate vote?

• Auckland Central (Tizard) by 160 votes
• Hamilton West (Gallagher) by 3,501 votes
• New Plymouth (Duynhoven) by 1 (yes 1) vote
• Otaki (Hughes) by 5,155 votes
• Palmerston North by 506 votes
• Rotorua (Chadwick) by 4,558 votes
• Taupo (Burton) by 5,700 votes
• West Coast-Tasman (O’Connor) by 3,274 votes

That would reduce Labour to 23 seats. But a Marae Digi-poll out on Sunday had the Maori Party ahead in all seven Maori seats. If they pick up the other three, then Labour would be down to just 20 electorate seats out of 70.

So is that the worse case scenario? Well no. This is based on a poll which has National ahead of Labour by just 13% in the EV. The previous CB poll had Nat 55% EV and Lab 33%. If that (unlikely) result had a linear application, how many more seats would fall:

• Hutt South (Mallard)
• Mana (Laban)
• Maungakiekie
• Port Hills (Dyson)
• Waimakariri (Cosgrove)

Also Rimutaka and Waitakere (Pillay) get reduced to under 500 vote majorities for Labour. If they also got taken on a bad swing, Labour would be reduced to 13 electorate seats.

Now these are not predictions by any means – it is just a linear application of poll results. There is no way anyone will be winning Otaki by 5,000 votes. I expect Nathan Guy and Darren Hughes to have a very close contest. They are both good MPs. The retirement of Steve Maharey might make Palmerston North very interesting.

Labour will have an interesting challenge with their list rankings. They want to rejuvenate and have some new MPs after the election. So one might expect they won’t automatically put sitting MPs at the top of the list. However if they do not protect sitting MPs, then quite a few of them could end up out of Parliament.