2026 New Zealand election forecast

A seat projection with visible uncertainty: baseline Parliament, sensitivity to a two-point party swing, the 5% threshold and the coalitions each scenario permits.

Mechanical estimate · Average v1.0 · 1 September 2026

Opposition bloc56LAB + GRN + TPM

Majority66of 130 seats

Government bloc65NAT + ACT + NZF

LAB 35GRN 15TPM 6OPP 9NZF 12ACT 10NAT 43

NAT4330.3%LAB3529.1%GRN1511.9%NZF129.7%ACT108.6%OPP97.2%TPM62.0%

Model sensitivity

How much can two points change?

National's range is constrained by the model's current 43-electorate assumption; even a lower party vote cannot reduce it beneath seats already assumed won. Each row otherwise moves one party two percentage points down and up while proportionally adjusting the others. These are stress tests, not confidence intervals.

Party	Current vote	Baseline seats	Range
National Party	30.3%	43	43–43
Labour Party	29.1%	35	33–38
Green Party	11.9%	15	12–17
ACT New Zealand	8.6%	10	8–13
New Zealand First	9.7%	12	9–14
Te Pāti Māori	2.0%	6	6–6
Opportunity Party	7.2%	9	6–11

Threshold effect

Why 5% changes the whole Parliament

The same party moving only 0.2 points across the threshold can redistribute seats across every eligible party.

4.9% OPP

Opportunity below threshold

0 seats

Opportunity receives no list seats under this assumption.

LAB 38GRN 16TPM 6NZF 13ACT 11NAT 43

5.1% OPP

Opportunity clears threshold

6 seats

Opportunity enters Parliament with 6 estimated seats.

LAB 36GRN 15TPM 6OPP 6NZF 12ACT 11NAT 43

Government formation

Coalitions under the baseline

The baseline is one point estimate. This table shows which familiar combinations cross the majority line before any negotiations or support agreements.

Build a custom scenario →

NAT + ACT + NZF651 shortNational + ACT + NZ First

LAB + GRN + TPM5610 shortLabour + Green + Te Pāti Māori

LAB + GRN + OPP597 shortLabour + Green + Opportunity

NAT + ACT + OPP624 shortNational + ACT + Opportunity

What this forecast does — and does not do

It translates the NZ Polls average into Parliament using Sainte-Laguë allocation, the 5% threshold and explicit electorate-seat assumptions. It does not estimate win probabilities, forecast individual electorates or pretend poll error is independent between parties.

Read the full model methodology or change the vote shares yourself.