New Zealand Election 2017

[Polkergeist]

What are the odds of Winston Peters holding Northland?

[Polkergeist]

Well with both Greens and NZF flirting with 5 % at some point this campaign, its not out of realm of possibilities that both miss out on seats.

Either of the majors could win a majority too...

[Polkergeist]

The march of advance voting continues unabated

http://www.elections.org.nz/events/2017-general-election/advance-voting-statistics

Ordinary and special votes cast within New Zealand - 9 days before election

2011:   57,347
2014:   147,560
2017:   229,259

Keep in mind there was no voting in the third last week this time

[Polkergeist]

Ordinary and special votes cast within New Zealand - Sunday before election

2011:   90,861 (27% of final advance votes cast)
2014:   225,513 (31% of final advance votes cast)
2017:   445,350

http://www.elections.org.nz/events/2017-general-election/advance-voting-statistics

While there are some differences between 2017 & the past 2 elections, the majority of the advance votes are cast in the last 5 days.

If we assume that 70% of advance votes are cast in the last 5 days, the number of advance votes cast could be about 1.1 million.

Assuming an 80% turnout on a roll of about 3.2 million, then 1.1 million advance votes equates to about 40-45% of all votes cast.

All in all it will diminish the effect of the late swing on the final outcome.

[Polkergeist]

I still don't see how the nationals have shot up when nothing major has happened that would suggest something like this.

It has been 7 weeks since Ardern became leader, maybe the new leader poll bounce is wearing off.

[Polkergeist]

Also interesting both NZF and Green still flirting with 5%. While majority of polling evidence has both at 5% or over, still puts their place in Parliament in play.

Interesting thought, does Labour want NZF in Parliament?

If we take the CB as the current state of play, if NZF are knocked out, it turns Parliament into a straight National-ACT vs Labour-Greens-Maori two horse race.

If NZF are knocked out then Nationals could win outright or with ACT which is more or less the same...

[Polkergeist]

I still don't see how the nationals have shot up when nothing major has happened that would suggest something like this.

It has been 7 weeks since Ardern became leader, maybe the new leader poll bounce is wearing off.

Maybe, but if that was happening, your party wouldn't lose 6 points in one polling period, it would be sliding off a little bit slower, don't you think?

The previous poll could have been a labour friendly sample and the current poll a National friendly sample the underlying shift could be less than 6 points.

But even with the similar NewsHub poll tonight I would not rule out a Labour vote in the 40s. The polls have been bouncing around and we don't have too many data points

[Polkergeist]

The result of the advance votes at the 2014 election had very little difference to the overall result.

With a much larger advance vote this time we should see these small differences disappear....

Or should we? 

This time around we are seeing a much more energised left-of-centre voters.

Is it possible that the advance votes could lean to the left by a few points this time compared to the overall result?

[Polkergeist]

The result of the advance votes at the 2014 election had very little difference to the overall result.

With a much larger advance vote this time we should see these small differences disappear....

Or should we? 

This time around we are seeing a much more energised left-of-centre voters.

Is it possible that the advance votes could lean to the left by a few points this time compared to the overall result?

This could happen, and it's one of the reasons why I won't be relying too much on the final election day vote total, especially seeing how close it could be.

Yes, the special/overseas votes can always shift one seat from right to left. Have they ever moved more than 1?

[Polkergeist]

Final advance voting statistics are in.

http://www.elections.org.nz/events/2017-general-election/advance-voting-statistics

Total votes cast: 1,240,740
Voters on roll: 3,252,269
Advance votes cast as share of total roll: 38.15%

I expect given the tight contest that total turnout of enrolled voters should be 80%.

So advance voting has come close to but not exceeded 50% of all votes cast for the election

In terms of advance votes counted for tonight, I understand that special votes (people who enrolled when voting) will not be counted. 

At previous elections 85-87% of advance votes were able to be counted for the night, so I'd expect an advance count of about 1.05 million or about 40% of votes cast at the election

[Polkergeist]

To provide a bit of context to the early advance vote results I will be posting projections of the final result based on the early advance vote.

I have put together a model based on an adjusted advance vote for each electorate, a 50-50 combination of the 2014 advance vote and the 2014 non-advance vote. The adjustment is designed to reflect the increased advance vote at this election which has effectively doubled. 

The 2017 advance vote will be compared with the combined adjusted advance vote of the electorates which have reported. From this we can get a swing to apply to the final 2014 results to give a projection of the final 2017 result.

These are my projections and should not be confused with the official results of media projections.
 
If for some other reason the advance vote is different to the overall results (energised left-leaning voters, late swing), then the projection will reflect this and will be off from the final results

Even though at the last election the advance vote was close to the final result, in a close result I would wait until election day votes are in by sufficient numbers to call the result.

Based on 2014, this projection should be of most use the first hour of the count.

I hope this is of use to readers.

[Polkergeist]

Usually TVNZ streams its coverage and I have viewed it in Australia in the past.

[Polkergeist]

Polkergeist Projection

NAT   46.3%   56
LAB   36.5%   45
GRN   8.4%   10
NZF   6.3%   8
MRI   0.4%   2
ACT   0.4%   1
OTH   2.0%   0

Advance vote reported in 26 of 71 seats      

[Polkergeist]

Polkergeist Projection #2

NAT   46.5%   57
LAB   36.8%   45
GRN   5.7%   7
NZF   7.2%   9
MRI   0.6%   2
ACT   0.5%   1
OTH   2.7%   0

Advance vote reported in 48 of 71 seats   

I'm with Tender, the final polls are pretty much it at the moment, although NZF is outpolling the Greens

[Polkergeist]

Winston is leading in Northland

http://electionresults.govt.nz/electorate-details-35.html

[Polkergeist]

Also Labour is leading in all 7 of the Maori seats, even in the one seat the Maori Party holds Waiariki

http://electionresults.govt.nz/electorate-details-71.html

[Polkergeist]

The raw figures are from http://www.elections.org.nz/events/2017-general-election/advance-voting-statistics

The 80% turnout was my estimate.

[Polkergeist]

At this point the raw figures are as good as any as all seats have reported.

http://electionresults.govt.nz/index.html

Currently in terms of seats NAT 57 LAB 45 NZF 9 Grn 8 ACT 1

It's looking like Maori party are out. Their closest seat is Te Tai Hauāuru where they are running Rugby League player Howie Tamati

Winston has fallen behind National in Northland

[Polkergeist]

Matt McCarten just said Maori Party was out, but the panel is hailing his "bias". So I don't know what to make of that...

Even if Maori Party pull it out of the fire and win two seats. They can only be of help to National to form a Government if National win two more party seats than they are right now.

That would require a lift in the National party vote from here to about 47.5% which is a bit of an ask.

[Polkergeist]

National flirting with a 58th seat at the expense of Labour or the Greens with the first election day votes coming in from small mostly rural booths. I wouldn't call a significant difference in election day votes vs advance votes just yet.

[Polkergeist]

Nats have hit 59 seats on the raw results but One News says their computer projection has National on 56 seats and Labour on 46 seats.

At the moment the election day votes (about 330k)  are Nat 49.7% Lab 32.4%

[Polkergeist]

Nats back to 58 seats on the raw vote

Election day votes have started to drift back to advance votes/polls levels,

At 546k ed votes, Nat are 48.5%, Lab at 33.3%, NZF at 7.9%, Grn at 5.4%

Interestingly, the Opportunities Party is at 2.5% on ed votes.

[Polkergeist]

With about 950k advance votes counted and 700k election day votes counted, here is the comparison of the advance vs election day vote shares for each party

(advance vs election day)

National (45.6% vs 47.7%)

Labour (36.7% vs 34.0%)

NZF (7.1% vs 7.9%)

Greens (6.2% vs 5.4%)

[Polkergeist]

Now with 900k election day votes counted

(advance vs election day)

National (45.5% vs 47.4%)

Labour (36.7% vs 34.3%)

NZF (7.1% vs 7.8%)

Greens (6.2% vs 5.5%)

[Polkergeist]

I don't know about a landslide for National they haven't won a majority in their own right.

Labour seems to have taken most of their gains from the Greens. Ultimately Jacindamentum was a left-facing phenomenon.

Back to turnout, the political scientist on One just said turnout had gone into the low 80's.

I guess what is remarkable about Nationals' vote tonight is that they kept a relatively high vote share when turnout rose.