National has recorded its lowest party vote in the RNZ-Reid Research poll since Christopher Luxon became leader, falling 2.1 points to 28.7 percent. It is the party's fourth consecutive decline in the series. The poll, conducted between 2 and 9 July and published on RNZ's website on Tuesday, was carried out after the annual conferences of National, Labour and ACT and a run of policy announcements from all parties RNZ.

Labour led the party vote at 34 percent, down 1.6 points on the March 2025 poll. New Zealand First rose 0.9 points to 11.5 percent, its best Reid Research result since July 2017. The Greens edged up 0.2 points to 10.3 percent, and ACT gained 0.8 points to 7.8 percent. Te Pāti Māori sat on 2.3 percent, down 0.9 points.

The standout mover is the Opportunity Party (TOP), which more than doubled its support, up 2.7 points to 4.7 percent — 0.3 points shy of the 5 percent threshold required to enter Parliament without an electorate seat. Undecided or non-voters made up 3.8 percent of respondents, down sharply from 7.1 percent in March.

On these numbers, the governing coalition of National, NZ First and ACT would hold 61 seats in a 120-seat Parliament (National 36, NZ First 15, ACT 10), against 59 for a Labour-Greens-Te Pāti Māori bloc (Labour 43, Greens 13, Te Pāti Māori 3). That's a bare majority, and one sensitive to seat-count effects that don't show up in the topline party vote.

Two of those effects are worth setting out for anyone modelling the numbers through to an election. National currently holds 43 electorate seats. If it won more than 36 electorates on the night, it would create an overhang, adding seats to Parliament rather than taking them from coalition partners, and widening the coalition's effective margin. Te Pāti Māori, which won six of the seven Māori electorates last election, would need more than three electorate wins to start closing the gap with the government bloc even on a reduced party vote share. And should TOP clear 5 percent, the reallocation of list seats would reshape the numbers for every other party currently in Parliament, not just add a new one alongside them.

On preferred prime minister, Chris Hipkins remains in front at 23.9 percent, up 3.2 points. Luxon rose 2.7 points to 20 percent. Winston Peters was third on 13.2 percent, up 0.1. Fifteen percent of respondents either declined to name a preferred prime minister or said they didn't know. Both leaders posted their lowest-yet approval ratings in this poll, according to RNZ, though Hipkins' preferred PM figure moved up rather than down — the two measures aren't tracking together this round.

On the direction-of-the-country question, 34.7 percent said things were heading the right way, up 2.4 points, against 46.5 percent who disagreed, down 3.5 points — a net score of -11.8. A further 17.5 percent were undecided and 1.2 percent didn't know.

Set against other recent polling, the RNZ-Reid Research result sits in a slightly different place to Roy Morgan, whose June reading had the National-led government's combined support broadly steady at 51 percent Roy Morgan. The 1News Verian poll from late June had both major parties at 30-year lows, with Labour down five points on that survey and TOP also closing on the 5 percent line 1News. The consistent thread across the three pollsters, whatever the house-effect differences in absolute numbers, is a minor party squeeze on both flanks and a TOP vote that keeps building.

What that means for governing arithmetic is the more interesting question for anyone in the Beehive or the gallery right now. A coalition majority of two seats, on paper, leaves very little room for a byelection loss, a defection, or an electorate-seat surprise to flip the numbers. It also means the overhang scenario around National's electorate seats stops being a technical curiosity and starts being central to how any post-election negotiation would actually play out, because it changes the size of Parliament itself rather than simply the shares within it.

The TOP trajectory is the other thread worth watching closely between now and polling day. A party sitting 0.3 points under threshold, having more than doubled its support in a single poll, is in a different position to one languishing at 2 or 3 percent — the swing needed to cross the line is now small enough that a single further movement of this size would put list seats in play, with consequences for every other party currently banking on the status quo distribution.

None of this settles a widely used but imprecise piece of shorthand about where the parties actually sit going into the next campaign. Reid Research, Roy Morgan and 1News Verian use different methodologies and sampling windows, and their numbers won't ever align perfectly. But the direction of travel — a weakening National vote, a strengthening NZ First, and a TOP surge nobody had fully priced in a year ago — is now showing up across all three.