Is the Election Really That Close?

There have been two polls released in the last week, taken at the same time but showing two very different results. 1News-Verian showed National had dropped back below 30 per cent but the correlating gains for ACT and NZ First meant no change to support for the coalition overall. Contrarily, the RNZ-Reid Research poll suggests National’s support has collapsed below 2020 levels, providing a comfortable majority for a Labour/Greens/Opportunity Government.

RNZ-Reid Research is the left-wing version of the Roy Morgan poll: it tends to produce projections for the left above the general trend, while Roy Morgan tends to do the same for National. In the 2023 election, Roy Morgan was the least accurate poll compared to the final result, with RNZ-Reid Research being the second-least accurate poll.

With an overhang in seats for National and Te Pāti Māori being a genuinely accepted expectation for most commentators, that which once destroyed a party’s electoral prospects is now good news. Last week was visually horrific for Labour, following the resignation of Tāmaki Makaurau candidate Kerrin Leoni after photos emerged of a campaign event that looked too much like campaigning for gang member votes. Ironically, her resignation will actually help Labour form a government as Tāmaki Makaurau switches from a seat Labour probably would have won, to a seat which Te Pāti Māori will definitely win. This ensures Te Pāti Māori holds four electorates, instead of three, with their party vote sitting at 1.02 per cent they can expect an overhang of three seats.

Likewise, National’s 24.8 per cent in Thursday’s RNZ-Reid Research poll is good news for the prospects of the coalition being re-elected. Prior to the poll, National’s party vote indicated they could win 35 seats, while winning up to 39 electorates, to add four overhang seats in a 127-seat parliament. Now the Poll of Polls shows National only winning enough party votes for 34 seats, increasing their overhang to five seats in a 128-seat parliament, if they hold 39 electorates.

However, the election is still as close as it has been throughout the entire campaign. When the maximum likely overhangs are taken into account, National/ACT/NZ First has 64 seats out of 128 and Labour/Greens/Opportunity/Te Pāti Māori also have 64 seats out of 128. Te Pāti Māori’s support in their four seats is also much more solid and predictable than National’s support in 39. 

(The Poll of Polls includes three 1News-Verian polls, The Post/Freshwater poll and the RNZ-Reid Research poll covering the period of 13 October to 28 October 2026). 

●     Scenario 1 assumes no overhang.
●     Scenario 2 assumes all parties currently in parliament will hold at least one electorate seat.
●     Scenario 3 assumes National holds 39 electorate seats with no other party achieving an overhang.
●     Scenario 4 assumes Te Pāti Māori holds four electorate seats with no other party achieving an overhang (up from three, following the dis-endorsement of Kerrin Leoni by Labour in Tāmaki Makaurau).
●     Scenario 5 assumes National holds 39 electorate seats and Te Pāti Māori holds four electorate seats. 

Prime Minister Christopher Luxon has always insisted that he doesn’t talk about polls, but has decided to talk about the RNZ-Reid Research poll stating that there would be “dire consequences for Kiwis” if the poll were to be replicated as the election result. Given how close the contest is, should Luxon be regretting his decision to rule out forming a coalition with Opportunity Party?

Richard Prebble made an interesting observation, in his Substack I’ve Been Thinking, on 20 October. National uses British pollster Sancrox Political Advisory for their internal party polling and therefore will be privy to information that isn’t revealed by any New Zealand pollsters. What we do know from New Zealand pollsters indicates that Luxon’s continued refusal to work with Opportunity Party could be driven by indications that Opportunity’s support is too soft to be sure they will achieve five per cent and, by refusing to engage with them, National could ensure Opportunity doesn’t achieve five per cent of the vote.

Talbot Mills polling from July showed that 23 per cent of Opportunity Party supporters had voted for National in 2023. If Sancrox is running focus groups on current Opportunity supporters that asks whether they would vote Opportunity if there was no chance of them going into coalition with National, it is possible that most of those Opportunity Party supporters might decide to vote with the centre-right.

The only pollster that breaks down party support by demographic group is Roy Morgan. While the company is notorious for rogue polling, the breakdown of demographic support is still interesting. Roy Morgan’s September poll showed Opportunity Party had 6.5 per cent support overall. Amongst women that support is three per cent, with 2.5 per cent support amongst women 18–49 and 3.5 per cent support amongst women aged over 50. The party’s support amongst men was 10 per cent, with 13 per cent from men aged 18–49 and 6.5 per cent of men aged over 50. Young men are the least likely demographic to actually cast a vote in New Zealand and their proportion of Opportunity Party’s support is enough to sink the party’s vote below five per cent if they don’t turn out to vote.

Perhaps, it is worth adding a sixth scenario to the projected election results: Opportunity Party only achieving 4.9 per cent of the vote. Without any overhangs, National/ACT/NZ First has 63 seats vs Labour/Greens/Te Pāti Māori on 57 seats. Assuming Te Pāti Māori wins four electorates, National/ACT/NZ First would have 63 seats vs Labour/Greens/Te Pāti Māori on 60 seats. Assuming 39 electorates for National and four electorates for Te Pāti Māori, National/ACT/NZ First would have 66 seats vs Labour/Greens/Te Pāti Maori on 60 seats.

Perhaps the election isn’t as close as we think.