2017 General Election – Official Result

As predicted National have slipped after special votes have been counted, but perhaps more than expected, by 1.6% to 44.4%. They have dropped two seats to 56.

Both Greens and Labour have picked up a seat each, with a combined total now of 54 seats, still short of National.

Party Votes % of Votes Electorate Seats List Seats Total seats
National Party 1,152,075 44.4% 41 15 56
Labour Party 956,184 36.9% 29 17 46
New Zealand First Party 186,706 7.2% 9 9
Green Party 162,443 6.3% 8 8
ACT New Zealand 13,075 0.5% 1 1
The Opportunities Party (TOP) 63,261 2.4%
Māori Party 30,580 1.2%
Aotearoa Legalise Cannabis Party 8,075 0.3%
Conservative 6,253 0.2%
MANA 3,642 0.1%
Ban1080 3,005 0.1%
New Zealand People’s Party 1,890 0.1%
United Future 1,782 0.1%
NZ Outdoors Party 1,620 0.1%
Democrats for Social Credit 806 0.0%
Internet Party 499 0.0%
Total 2,591,896 71 49 120


We are in for another round of speculation and positioning as the negotiations get under way.

Here are the interim results for comparison:

Party Votes % of Votes Electorate Seats List Seats Total seats
National Party 998,813 46.0 41 17 58
Labour Party 776,556 35.8 29 16 45
New Zealand First Party 162,988 7.5 9 9
Green Party 126,995 5.9 7 7
ACT New Zealand 10,959 0.5 1 1
The Opportunities Party (TOP) 48,018 2.2
Māori Party 23,456 1.1
Aotearoa Legalise Cannabis Party 5,853 0.3
Conservative 5,318 0.2
MANA 2,775 0.1
Ban1080 2,440 0.1
New Zealand People’s Party 1,631 0.1
United Future 1,471 0.1
NZ Outdoors Party 1,333 0.1
Democrats for Social Credit 732 0.0
Internet Party 464 0.0
Total 2,169,802 71 49 120



Revised election statistics, turnout down

The Electoral commission has released revised election statistics.

  • Estimated eligible population: 3,569,830
  • Total number enrolled: 3,298,009
  • Election night votes counted: 2,169,802
  • Special votes still to be counted: 384,072
  • Total estimated votes: 2,563,740
    (Total 2014 votes counted: 2,416,479)

This has brought the turnout down to just below the turnout in 2014 (by my calculation).

  • Estimated 2017 turnout (of enrolled voters): 77.6%
  • Actual 2014 turnout (of enrolled voters): 77.9%

So that is a slightly lower turnout, despite the large increase in advance votes  and despite the claims of ‘youthquakes’.

  • Estimated 2017 turnout (of eligible voters):71.7%

This turnout based on eligible voters is not usually stated but I would have thought it more pertinent.







Caretaker government

New Zealand is operating under a caretaker government until a new government is formed.

The 51st Parliament ended on 22 August.  The 52nd Parliament must meet no later than 6 weeks after the ‘return of the writ’, due 12 October (unless a Judicial Recount is being conducted).

When will Parliament open again?

Before the 52nd Parliament can be formally opened, there are some important constitutional and practical things that need to happen.

The period after the election, while the government is being formed, is known as the “caretaker period.”

Under New Zealand’s proportional representation (MMP) electoral system, it is likely that two or more parties will negotiate coalition or support agreements so that a government can be formed, either as a majority or minority government. This can take a bit of time, but Parliament must meet on or before 23 November (two months after the election).

It is expected that we will have a new government worked out by then.

But note that Germany also had an election last weekend and Angela Merkel said she was confident of having a new government by Christmas: “I’m generally always confident. And for many years, I’ve gone by the motto: ‘power lies in tranquility.’”

In 2010-2011 Belgium went without an elected Government for 589 days because opposing parties were unable to agree on policy issues and form a governing coalition following elections.

It is unlikely to take this long here, but until a new government is formed we are operating under a caretaker government.

What is the caretaker period?

There must always be a Government, but during the period when the government is being formed, it’s in a “caretaker” mode. This means that, by convention, the Government that held office before the election will generally hold off on making significant decisions, new policy, or decisions with long-term implications. The exception to this is when there is an emergency or crisis.

Here’s how it works:

After the election on 23 September, the Prime Minister indicated that the Government that held office before the election, will operate in accordance with this caretaker convention. This continues until the political situation is resolved and a new Government has been sworn in.

Current Ministers continue with their existing responsibilities after the election, until new Ministerial appointments are made or their responsibilities are reassigned. Ministers who are not returned as MPs may continue in office as caretaker Ministers for a period, but must leave office no later than 28 days after polling day (that is, by 21 October 2017).

So Peter Dunne must still be a caretaker minister even though he didn’t stand for ere-election. And also Te Ururoa Flavell, who lost his seat in the election.

There are two main scenarios for caretaker government.

  1. Where it is not clear who will form the next government

The normal business of government, and the day to day administration of departments and other agencies in the State sector continues as usual.

Decisions taken and specific policy determined before the start of the caretaker period may usually be implemented.

Decisions on significant issues, new policy or changes to existing policy, and issues with long-term implications should be deferred if possible. If deferral is not possible, short-term solutions should be sought. If this is not feasible, decisions should be made after consultation with other parties.

  1. Where it is clear who will form the new government, but they have not yet taken office

The Government continues in caretaker mode until Ministers are formally appointed. The outgoing government should undertake no new policy initiatives, and should act on the advice of the incoming government on any significant constitutional, economic or other issue that cannot be delayed until the new government formally takes office – even if the outgoing government disagrees with the course of action proposed.

So the returning MPs who are ministers will also be caretakers – if National doesn’t do a governing deal with NZ First they will be acting ministers until new ones are sworn in, but with limited duties.

Source: Parliamentary Service

No new legislation can be passed until the new Government takes over.

Would a lower threshold have made much difference?

The large parties have kept ensuring it remains very difficult for small parties to succeed or survive by keeping a ridiculously high threshold of 5%. A slightly more reasonable 4% was recommended, and many people have said it should be much lower.

No new parties that didn’t have MPs who had jumped ship have succeeded in getting into Parliament in 21 years of MMP.

What would the election result have looked like with no threshold? From Rediit What Parliament could have been if there was no 5% minimum

Actual seats (provisional):


Seats with no 5% threshold (provisional)



That would have brought 5 MPs from 3 parties into Parliament that didn’t make it.

This wouldn’t have changed the National+NZ First majority much (58 to 54, still a clear majority) but it would have left Labour+NZ First+Green short on 59 (they have a bare majority 61 seats now).

However also noted at Reddit:

Of course, if there was no 5% minimum, peoples voting behavior would have changed.

I suspect greens would have gotten less, with TOP and maybe United Future getting more.

And if we had also had no threshold last election the Conservatives and Internet-Mana may have been in Parliament and in the mix this election. Even a slightly lower 4% could have made a significant difference.

Keeping the 5% threshold in place is keeping new parties out of Parliament and gradually squeezing small parties out too. At one stage polls suggested that both NZ First and Greens were at risk of missing the cut.

Thresholds in other countries:

  • Germany 5%
  • Poland 5%
  • Israel 3.25% (it has gradually been increased)
  • Turkey 10%
  • Netherlands effectively 0.67%
  • Slovenia effectively 4%
  • Sweden 4%

The Parliamentary Assembly of the Council of Europe recommends for parliamentary elections a threshold not higher than 3%.

There are some variations. In New Zealand if a party wins an electorate seat the threshold doesn’t apply to them, in Germany they have a 3 seat exemption.

But this is probably all futile pondering, National and Labour seem intent on excluding fair representation by maintaining a high 5% barrier to protect their own interests at the expense of  fair democracy.

How will special vote affect the result?

The short answer is that no one knows, and there’s no way of knowing until the final results are published on 7 October (Saturday next week).

But that hasn’t stopped people from guessing, or assuming based on past effects of special votes on election night totals.

It is widely presumed that National will lose a seat or two (they lost 1 in 2014 after specials were counted) or possibly two, and Greens are expected to pick one up and possibly Labour as well.

But this election has been quite different to past elections.

There are an estimated 384,072 special votes to be counted, up from about 300 thousand in 2014 and about 15% of the total vote. They include:

  • votes cast overseas;
  • votes cast by the telephone dictation service;
  • votes cast on polling day by people voting at a voting place not designated to serve their electorate;
  • votes cast by people who enrolled after the printed electoral roll was closed (including during advance voting);
  • votes cast by people on the unpublished roll; and
  • votes cast by people who think they’re enrolled to vote, but aren’t (these votes don’t count).

It is assumed that many young people enrolled and voted, especially at university polling places, so that is likely to favour Labour and the Greens. But the turnout of younger voters is usually relatively low.

In Election 2017: the Special Votes Graeme Edgeler tries to estimate the shift in support from specials based on the movement last election.

…using the same rudimentary method I used last time (assuming the variance in special votes is the same size as it was at the 2014 election), along with the Electoral Commission’s estimate of the number of special votes at this election, I predict the following final result:

  Preliminary Projected
  Vote share Seats Vote share Seats
National 46.03% 58 44.86% 56
Labour 35.79% 45 36.55% 46
New Zealand First 7.51% 9 7.32% 9
The Greens 5.85% 7 6.32% 8
TOP 2.21% 0 2.35% 0
Maori Party 1.08% 0 1.12% 0
ACT 0.51% 1 0.51% 1

In 2014, National did 17% worse on special votes than they did with ordinary votes, while the Greens did 53% better. This was enough to see National lose one seat after special votes were counted, and the Greens to pick one up.

Edgeler makes it clear that this is a guess:

This time, assuming (perhaps foolishly) that the same basic numbers apply, and with the larger number of Special Votes still to be counted, both Labour and the Greens are within striking distance of of taking a list seat from National.

There is something else we may get an idea from – the trends in polls during the  voting period (advance votes and election day).

We don’t know when Special Votes were cast. We can assume that many were cast during advance voting, but there could also have been a lot on election day.

The voting period was from 11 September to 23 September. Most of the advance votes were in the last few days, not covered by polling periods.

Polls show that before voting started in early September they favoured Labour over National, and Greens had already recovered after earlier bottoming out below 5%.

The last polls before the election:

  • Colmar Brunton 2-6 September: NAT 39, LAB 43, NZF 9, GRN 5
  • Roy Morgan 28 Aug-10 Sep: NAT 40, LAB 39.5, NZF 6, GRN 9
  • Reid Research 6-11 September: NAT 47.3, LAB 37.8, NZF 6, GRN 4.9
  • Colmar Brunton 9-13 September: NAT 40, LAB 44, NZF 6, GRN 7
  • Reid Research 13-20 September: NAT 45.8, LAB 37.3, NZF 7.1, GRN 7.1
  • Colmar Brunton 15-19 September: NAT 46, LAB 37, NZF 4.9, GRN 8

The last polls have similar results to the election, suggesting the poll swings had settled down by the second week of voting when most advance votes were cast.

This is inconclusive. Reid Research suggests support had already swung back to National from Labour while Colmar Brunton suggests a late swing.

Both show a late swing to the Greens but mostly before voting started. NZ First were up and down.

There is really not enough information about the special votes to do anything other than guess what effect they will have on the outcome.

All we really know is that there was quite a bit of late support movement , and this may or may not impact on special vote support of various parties.

One notable thing about voting patterns on election night – pundits kept suggesting, based on past election patterns, that as the results came in support was likely to swing against National by a percent or two (44% was mentioned) and increase for Labour. This barely happened – National support dropped a little but stayed in the 49-47% range.

But we simply don’t know if a smaller swing against National than last election is likely or not after special votes are counted.

We can guess, but we should be careful about having much confidence in assumptions that pundits make. There have also been many surprises in this election.

Total votes and turnout

Total provisional votes: 2,169,802

Special votes (about): 384,000

Estimated total votes: 2,553,802

Total votes in 2014: 2,416,479

Approximate increase in votes 2014 to 2017:  137,323

Estimated eligible voters: 3,569,830

Estimated turnout of eligible voters: 71.5%
BUT the statistics commonly used are percentage of enrolled voters.

Estimated turnout of enrolled voters: 78.8%

So that is a slight increase on last election (where Labour did very poorly). Past turnout:






Election – provisional results

The 2017 election night provisional results (which excludes hundreds of thousands of special votes):

Party Votes % of Votes Electorate Seats List Seats Total seats
National Party 998,813 46.0 41 17 58
Labour Party 776,556 35.8 29 16 45
New Zealand First Party 162,988 7.5 9 9
Green Party 126,995 5.9 7 7
ACT New Zealand 10,959 0.5 1 1
The Opportunities Party (TOP) 48,018 2.2
Māori Party 23,456 1.1
Aotearoa Legalise Cannabis Party 5,853 0.3
Conservative 5,318 0.2
MANA 2,775 0.1
Ban1080 2,440 0.1
New Zealand People’s Party 1,631 0.1
United Future 1,471 0.1
NZ Outdoors Party 1,333 0.1
Democrats for Social Credit 732 0.0
Internet Party 464 0.0
Total 2,169,802 71 49 120

‘Special votes’ also include those cast from outside an electorate, people who enrolled after writ day (August 23), those on the unpublished roll and those who cast special votes because of a complication that meant accessing a polling booth would cause hardship.

There were 303,000 special votes in 2012, I don’t have that number yet for this election.

This is the first time since 2002 that there are just 120 seats.

National 46.0% – this is a very good result for them for a fourth term government, but they are short of an easy majority due to a lack of support parties. It is a very good result for Bill English – If John Key hadn’t stepped down he would have struggled to do as well.

Labour 35.8% – 2 months ago they would have loved a result like this, 2 weeks ago they looked like they could do much better. Jacinda Ardern was disappointed they didn’t do better and looked like she lost last night. They could still form a government, and if not are at least on the road to recovery after a very poor 2014 election.

NZ First 7.5% – Winston Peters will be disappointed to have dropped 1.3% from their 2014 provisional vote. He claimed the polls were wrong but it turns out he was wrong. Still, it seems likely NZ First could decide who will lead the next government and are in a strong negotiating position, albeit with a smaller mandate.

Greens 5.9% – 2 months ago they would have seen a result like this as a disaster, they were aiming for three times as much, but beating the threshold is now seem as a major victory. James Shaw has done very well to lift them out of the Metiria mess, but it is still a poor mandate compared to their 2014 result. This is their worst result since they got 5.3% in 2005 (and were kept out of government by Winston Peters).

TOP 2.2% – an ok result for a new party but still well short of success, so it’s a failure for them.

Maori 1.1% –  down on their 1.32% in 2014, and out due to Te Ururoa Flavell losing his electorate. A disaster for the party, and if Labour don’t form the next government a poor result for Maori voters.

ACT Party 0.5% – down on their 0.69% in 2014 so failure as a party. David Seymour keeps them in Parliament through winning Epsom, but ACT have to get their act together, somehow.

United Future 0.1% – they were always going to struggle without Peter Dunne and they did struggle. Must now be history.

Conservative Party (0.2%), MANA (0.1%) and Internet Party (just 464 votes) – all dismal results.

Cannabis Party, Ban1080 Party, NZ People’s Party and NZ Outdoors Party were ‘up’ amongst the minnows but were ignored by media and largely ignored by voters.

Source: http://www.electionresults.govt.nz/

Comparing to last election – note that it is expected there could be similar movements in party support between provisional and final results, but there is no guarantee similar patterns will occur this year.

Table 2: 2014 Final Party Vote and Seat Count

Party Valid Party Votes
Provisional Total (1) Final Total Provisional Share (%) Final Share (%)
National Party 1,010,464 1,131,501 48.06 47.04
Labour Party 519,146 604,535 24.69 25.13
Green Party 210,764 257,359 10.02 10.7
New Zealand First 186,031 208,300 8.85 8.66
Māori Party 27,074 31,849 1.29 1.32
ACT New Zealand 14,510 16,689 0.69 0.69
United Future 4,533 5,286 0.22 0.22
Conservative 86,616 95,598 4.12 3.97
Internet-MANA 26,539 34,094 1.26 1.42
Other (2) 16,994 20,411 0.8 0.84
Total Valid Party Votes 2,102,671 2,405,622 100% 100%
Informal / Disallowed Votes (3) 9,851 40,675
TOTALS 2,112,522 2,446,297
  1. Provisional figures are as at election night on the 20th September; final figures as at 10 October, 2014. The preliminary results (announced on election night) include all ordinary votes – those cast by people who appear on the printed roll in the voting place they go to, and that are cast either on Election Day or in advance of Election Day. Provisional figures exclude the 329,726 special votes cast by voters who are overseas, or who are voting outside their electorate, or who are not on the printed roll for their electorate. Valid special votes are admitted to the final official count.
  2. Final party vote share for parties in the ‘Other’ category include: Aotearoa Legalise Cannabis Party (0.46%); Ban1080 (0.21%); Democrats for Social Credit (0.07%); The Civilian Party (0.05%); NZ Independent Coalition (0.04%); Focus New Zealand (0.03%).
  3. Informal Votes are votes cast by electors, qualified to vote, which are not counted because the ballot paper did not clearly indicate the party or the electorate candidate vote, or both. ‘Disallowed votes‘ are votes disallowed for one of the following reasons: no ground stated for a special vote, arrived late, declaration not enclosed, incomplete declaration, ballot paper not enclosed, or address invalid for electorate.

Source: Electoral Commission: http://www.electionresults.govt.nz/

UPDATE: The Electoral Commission estimates about 384,000, or 15 percent, of votes are special declaration votes and are still to be counted.

1.2 million advance votes

There have been 1,240,740 advance votes this election.

In 2014 there were 717,579 so there are nearly twice as many this time.

There were 2,416,479 votes in total in 2014, so there nearly 30% advance votes. This time it looks like being closer to 50% advance votes.


Advance votes that aren’t special votes are being counted from 9 am today. The Electoral Commission has a target of reporting them by 8.30 pm tonight, so there will be a big chunk of votes counted by then, possibly earlier.

That may or may not give us a good idea of what the final result may be. Polls were fluctuating a lot during the advance vote period (the last two weeks). If there was a late swing that won’t be evident until most votes are counted tonight, and then there will be a lot of uncounted special votes that could swing things again, further or back the other way.

Votes that were cast at the same time as enrolment are special votes so they won’t be counted today, they will be included in the final count on 7 October. Last election there were nearly 300,000 special votes.

There could be anywhere from 0.3 to half a million special votes this time, so if the result is close tonight, especially if any of Greens, NZ First or TOP are close to the threshold,  there could be a lot of uncertainty about the result for two weeks. If results in key electorates Northland and Epsom are close there will be further uncertainty.

We will find out a lot in a few hours time, but may have to be patient for a couple of weeks.

Daily totals for 2017 are approximate only and may be subject to change.

Date 2017 General Election 2014 General Election 2011 General Election
Wed-Sat 49497 18818
11 September 39570 22234 8893
12 September 48238 22846 8845
13 September 66084 23950 9750
14 September 72932 29033 11041
15 September 82422 31753 13399
16 September 89480 46200 20115
17 September 45306
18 September 105228 60966 27841
19 September 123307 62595 32937
20 September 133813 86021 38380
21 September 180887 122017 64137
22 September 253473 160467 80402
Overall total 1240740 717579 334558


How soon will we know the election result?

Due to large numbers of advance votes, which will be counted during the day tomorrow and will presumably be released soon after voting closes, we may get a good idea of the likely outcome soon after after 7 pm Saturday night.

But if the overall result is close, or if Greens, NZ First or TOP are close to the threshold, or if key electorates such as Northland or Epsom are close, then we may have to wait until later in the evening, or two weeks for official (final) results on 7 October.

Up to Wednesday 806,043 people had advance voted, more than the total  of 717,579 for 2014.

Cumulative Advance Votes Issued

With two days to add that looks likely to get well over a million early votes (I’ll update this post when yesterday’s numbers become available today).

It could get close to a half of votes being cast in advance this election. The total votes in 2014 were 2,416,479 – there may be more this time as we have a bigger voting population unless turnout drops.

Advance votes will be counted during the day tomorrow, so a large number of votes should be announced soon after voting closes at 7 pm tomorrow. We may pretty much know the election result fairly early.

A few people have been saying Don’t bother waiting up. Advance votes will tell you who won

As advance voting has increased advance voting totals have got closer to final results:

Party vote change 2002 to 2014 from advanced to final.png


The 2014 swing from National to Labour was close to 1.5%, and that swing the result (or at least the party with the biggest vote).

Greens improved by about 0.7% and that may swing them from under the threshold to over it.

If the total vote swings the same way for NZ First it could drop them below the threshold.

Will TOP defy polls and get close to The Opportunities Party?

There are also some key electorates. Northland will be crucial for NZ First if they miss the threshold. Epsom is crucial for ACT. The Maori Party need to win at least one electorate (they look odds on to win two).


Votes that are cast when people enrol at the same time are Special Votes, so are not counted on election night. There were nearly 300,000 special votes in 2014 that weren’t counted on election night.

There have been reports that many young people have been enrolling and voting. This could skew the special votes more in favour of Labour and the Greens.

Graeme Edgeler analysed Election 2014: The Special Votes

…adopting the same method I used last time, based solely on how special votes broke in 2011, along with the Electoral Commission’s estimate of the number of special votes at this election, I predict the following final result:

Preliminary Estimated
National 48.06% 61 47.24% 61
Labour 24.69% 32 25.05% 32
Green 10.02% 13 10.50% 13
New Zealand First 8.85% 11 8.52% 11
Māori Party 1.29% 2 1.38% 2
ACT 0.69% 1 0.68% 1
United Future 0.22% 1* 0.21% 1*
Conservative 4.12% 0 3.92% 0
Internet MANA 1.26% 0 1.38% 0
121 121

This underestimated changes in overall percentages. The final results and percentages:


If the advance voting totals are close, or if the election night votes are close, then we may have to wait another two weeks to find out the final result.

If any parties are in doubt due to close electorate or threshold results governing and coalition negotiations may not be able to seriously begin until after October 7 when final results will be announced.

And remember that the party with the most votes may not be able to form a government, so the advance vote ‘winner’ or the election night ‘winner’, may not end up being the winner by votes, and may or may not lead the next government.


This is the last day of the campaign!

Today is the last day of the election campaign for 2017. I’m sure many politicians, candidates, journalists and others such as myself will be relieved when it’s over.

All billboards and other public advertising must be removed by midnight and out of sight by election day (ridiculous given the number of advance votes these days), so campaign teams will be busy today cleaning up their efforts to entice voters.

If you want to promote any candidates or parties here get it done today, because you won’t be allowed to tomorrow because of electoral law.

I won’t be doing any posts on New Zealand politics or the election during the day on Saturday, to reduce the risk of anyone breaking the law and putting the site at legal risk.

I will put up an election night post for anyone who wants to comment after voting closes at 7 pm.