New Zealand 2026 General Election: electoral projection model

Final 2026 boundaries · notional 2023 results on new boundaries · national party-vote inputs drive logit, uniform or proportional swings + MMP

120-seat base Parliament · threshold + coattail + overhang

National party vote / primary vote

Change the national shares below. Each change is applied to every electorate as a logit, uniform or proportional swing (pick one below) and also feeds the MMP calculation.

Named parties 96.61% · Other 3.39%

National

Labour

Green

ACT

NZ First

Te Pāti Māori (TPM)

OPP

Other parties

3.39%

Find electorate

Highlights a seat on the cartogram.

Swing:

Polls used and weights

Download JSON

Use projected cartogram electorate winners

The seven named party-vote shares come from the main-page inputs. Turn off auto mode to manually set electorate-seat counts; manual electorate totals must equal exactly 71 or the output is blocked with an error. Custom parties below subdivide the main-page “Other parties” share.

Party	Party vote %	Effective vote %	Electorate seats	Threshold	Seat entitlement	List MPs	Overhang	Total MPs

How it works: a registered party qualifies if it gets at least 5% of the party vote or wins at least one electorate. Qualifying parties are allocated the available base seats with Sainte-Laguë (divisors 1, 3, 5, 7…). A party below 5% that wins an electorate therefore brings its whole party vote into the allocation, the electorate-seat exemption often called the “coattail” rule. If a party wins more electorate seats than its proportional entitlement, it keeps them and those extras become overhang seats; other parties keep their allocated seats. OPP is modelled separately from the residual Other parties share. The residual Other parties share is an aggregate for electorate swing. In the MMP tab you can subdivide that Other share into specific custom parties; any remainder is treated as scattered/non-qualifying vote. This lets you test another minor party crossing 5% or qualifying through an electorate win without changing OPP or the electorate model's residual Other swing.

Independent / unregistered electorate seats: In auto mode this follows any cartogram seat won by the aggregate “Other” category.

Methodology

Swing method:

Manual seats: click a party card in the seat-share panel to select it, then left click hexes to assign that winner. Click the selected party card again to clear your choice; with no party selected, right click a hex to reset it. Reset to non-manual edits clears all manual seat overrides but keeps the current polling swings; Reset to 2023 notional clears manual overrides and returns the national inputs to the 2023 baseline from any setting; 2023 notional ON and Poll average ON are greyed out while the page is on that setting. Yellow = projected gain (a different party from the 2023 incumbent wins), bright green = hold of a seat the 2026 boundaries made notionally another party's, orange = gain set by hand, pink = hold set by hand. Tāmaki Makaurau is set to Te Pāti Māori by hand by default: Labour's candidate resigned from the party after nominations closed (she stays on the ballot as Labour's), and the swing alone gives Labour the seat. Assigning it to another party on the map overrides that.

Overhang strategy: the button beside Reset to non-manual edits models split-ticket voting aimed at overhang. Only the candidate votes move: the party votes stay on the poll average, so the strategy wins electorates beyond what a party's list share covers without handing list seats to its partners. Both sides work the same way, the gaining party taking its candidate vote half from each partner's candidate. In each Māori electorate Te Pāti Māori gains 12 points of candidate vote, 6 from the Labour candidate and 6 from the Green. Tāmaki Makaurau goes to Te Pāti Māori by hand with or without the strategy (see the colours above); on the current poll average Labour still leads there after the 12-point swing, and a uniform boost big enough to take it would take Te Tai Tokerau as well. In each general electorate National gains 3 points, 1.5 from the ACT candidate and 1.5 from NZ First's. That size follows the polls: it is the smallest of 1, 1.25 or 1.5 points each that still gives the current government (National, ACT and NZ First) a majority on the poll average, overhang seats included, or 1.5 when none does, and it is worked out again from the latest poll average each time the button is clicked. Where a partner's candidate has less than its half (or none), the other partner's candidate covers the gap. Clicking it sets the party votes to the poll average and applies the strategy. The party-vote boxes can then be edited and the strategy stays on: the edited votes are the party vote, so they move the list seats, while every electorate stays as the poll average and the strategy make it (applying the poll average again refreshes that). National's size stays as it was picked. Clicking it again turns it off, and so does assigning a seat or editing the MMP table, since electorate votes cannot be edited by hand yet. Reset to 2023 notional turns it off too.

Electorates: 2026 structure: 48 North Island general electorates, 16 South Island general electorates and 7 Māori electorates = 71 electorate MPs. In a normal 120-seat Parliament this leaves 49 list seats before overhang. The 64 general electorates share one General block drawn in New Zealand's shape, with the South Island beside the Wellington end of the North Island; seats are placed so that real neighbours on the final August 2025 electorate maps touch where the hex grid allows. The Māori electorates box is a small New Zealand too: the six North Island seats above, Te Tai Tonga below-left like the South Island; Te Tai Tonga also includes part of the Wellington region and the Chatham Islands.

Baselines: Electorate baselines use Ben Raue/The Tally Room’s notional 2023 top two on the final 2026 boundaries and preserve his published margin. Lower-party shares come from the most-aligned predecessor’s 2023 candidate mix; 2023 Opportunity vote is kept separate from Other only where an Opportunity electorate candidate actually stood. Aligned incumbents remain separate for gain/hold highlighting. Port Waikato remains the actual 25 November 2023 by-election result. For lower-ranked candidate baselines, OPP represents the 2023 Opportunity candidate vote only in electorates where Opportunity actually fielded a candidate; all other minor/independent candidates remain in Other.

The compass. Scales (0–10), the same as the Latvia compass: Economic = left → right on the economy (state ownership, tax and spending, regulation, redistribution). Social = progressive → conservative (GAL–TAN: rights and lifestyle issues, immigration, law & order, nationalism, environment). Treaty = 0 Tiriti partnership / co-governance → 10 “one law for all”.

Pattern: Treaty 3.5 or less = Tiriti partnership (stripes), 6.5 or more = one law for all (checks), in between = plain.

How the scores are built: a 100-statement quiz on 2026 issues (50 economic, 50 social), with every party graded from strongly disagree to strongly agree on each statement. Parties that have governed nationally since 2016 (or Scotland, for the SNP) are graded on their record in office, not on promises that never passed: the Biden administration 2021–25, the Trump administrations, the Starmer, Albanese, Trudeau/Carney and Luxon governments, NZ Labour 2017–23, the UK Conservatives 2016–24 and the Australian Coalition 2016–22. Other parties are graded on their platform and parliamentary votes. Each axis score is the average over its 50 statements, shown as 2 + 0.8 × the raw quiz score: no real party gives the most radical answer to every statement, so the most left-wing parties sit around 2 rather than 0 and today’s centre-left governments sit near the centre. Order and relative distances are unchanged (raw scores are in the workbook). Economic topics: tax, spending, health, welfare, work, ownership and trade, housing, education. Social topics: immigration, gender & LGBT, life & religion, crime, climate, nation & culture (incl. Indigenous rights), democracy & speech. Hover a bubble or ring for its topic scores and grading basis; the full quiz and every grade are in the companion workbook.

Reading the US parties: on its record the Biden administration scores 6.0, centre-right: the most right-wing of the Anglosphere’s centre-left governments (Ardern/Hipkins 4.4, Albanese 5.0, Starmer 5.2, Trudeau/Carney 5.4) and about level with the vote-weighted centre of the UK, Australia, Canada and NZ (6.1). No universal health care, national paid leave, sectoral bargaining, federal minimum-wage rise or top-rate tax rise passed; its industrial policy, student-debt relief and drug-price caps did. On the platform it campaigned on it would score about 4.8.

Averages: NZ uses the seven named parties’ shares from the vote inputs; each comparison country uses its last national election (UK 2024 general, Australia 2025 House first preferences, Canada 2025 federal, US 2024 presidential). The Teals are independents, so they’re left out of Australia’s average. “All four” averages the four country averages equally.

Stances: under 3.5 left / progressive, 3.5–4.5 centre-left / lean progressive, 4.5–5.5 centre, 5.5–6.5 centre-right / lean conservative, over 6.5 right / conservative.

Compare with:

Tiriti partnershipone law for allin betweencomparison partyvote-weighted average partyBubble size = party vote from the inputs above; faded = under 5% with no electorate. Click a bubble to add it to the coalition.

Party	Economic	Social	Treaty

Treaty pattern: stripes = Tiriti partnership (3.5 or less), checks = one law for all (6.5 or more). The scales, how each party is graded, the US parties and the averages: Methodology tab.

Seats come from the MMP seats tab (your inputs, with any overhang) and run left → right in the same order as the coalition bar. The dashed line marks half-way. Click a party’s seats or its row to add it to the coalition, as on the coalition bar. Other (custom parties and independents) can’t be picked.