The Greens, Act New Zealand and Te Pāti Māori stayed fairly steady compared with the previous poll, while New Zealand First gained somewhat.
Party gains
- National: 37.6 percent, up 2.2 points (47 seats)
- Labour: 25.9 percent, down 3.5 (33 seats)
- Greens: 12.5 percent, down 0.2 (16 seats)
- Act: 9.1 percent, down 0.6 points (11 seats)
- NZ First: 7.3 percent, up 1.7 (9 seats)
- Te Pāti Māori: 3.5 percent, down 0.5 (6 seats)
National, Act and NZ First would still comfortably be able to form a government, with 67 seats putting them well above the 61-seat threshold in a 122-seat Parliament.
National would also still be two seats down compared with the election results, with NZ First up one. Labour would have one extra seat based on this poll, while the Greens would have one more - as well as the seat currently taken by now-independent MP Darleen Tana.
Tana could hold on to that seat and an MP's salary until the next election unless she resigns, the Greens use the party-hopping legislation to eject her, or she is convicted of a crime with a penalty of two years or more.
For minor parties, TOP was up 1.6 points to 2.4 percent, Outdoors & Freedom was down 0.3 points to 1 percent, and the combined total for all other parties was 0.8 percent.
National's leader and Prime Minister Christopher Luxon also made significant gains on preferred prime minister rankings, gaining 9.1 points to his highest rating of 34.5 percent.
Preferred prime minister:
- Christopher Luxon: 34.5 percent - up 9.1 points
- Chris Hipkins: 18.7 percent - down 1.5 points
- Chlöe Swarbrick: 10.9 percent, up 2.3 points
- Winston Peters: 5.6 percent, up 1.8 points
- David Seymour: 5.6 percent, up 0.6 points
- Jacinda Ardern: 5 percent, down 3.3 points
- Debbie Ngarewa-Packer 1.8 percent
- Chris Bishop 1.7 percent
- Nicola Willis: 1.2 percent
- Marama Davidson: 1.2 percent
- Leighton Baker: 0.4 percent
- Rawiri Waititi, 0.3 percent
The poll surveyed 1000 eligible voters and was weighted for demographics, with a margin of error of 3.1 percent at the 95 percent confidence interval. It was conducted between July 4 and 8.
Polls compare to the most recent poll by the same polling company, as different polls can use different methodologies. They are intended to track trends in voting preferences, showing a snapshot in time, rather than be a completely accurate predictor of the final election result.
The number of decided voters on the vote questions was 906. There were 34 (3.4 percent) undecided voters and 23 (2.3 percent) who refused the vote question.
