Can the left govern on 64 seats — or has the poll missed the overhang again?
- Grant McLachlan
- 15 hours ago
- 8 min read
The latest 1News Verian poll shows a left bloc able to form a government without New Zealand First. Two things it does not model — the overhang building in National’s electorates, and a rising Opportunity Party — could each hand the balance of power straight back to Winston Peters.
New Zealand watched the 6 o’clock news bulletin on 23 June to discover a poll that looked, on its face, decisive.
The 1News Verian poll taken between 13 and 17 June put both major parties at a thirty-year low, the combined National and Labour vote down to 61 per cent — the weakest since the first MMP election in 1996.
On the seat table that ran alongside it, a left bloc of Labour, the Greens and Te Pati Maori held 64 seats to the right bloc’s 60: enough to govern without Winston Peters.
It is a striking result.
It is also, as Verian states in its own poll report, a snapshot of party support and “not a prediction”.
That distinction matters more than it sounds.
The figure that turns party support into a government — the 124-seat table — rests on assumptions about an election the poll never measures: the contest in the electorates themselves.
Loosen two of those assumptions, both already visible in the same poll, and the identical party-vote numbers stop delivering a left government at all.
Contents
A snapshot, not a prediction
Verian’s seat projection is an honest piece of arithmetic, and it does not hide its workings.
The report states plainly that the calculation “assumes that Te Pati Maori hold their six electorate seats”, and applies the Sainte-Lague method the Electoral Commission uses to convert party votes into seats.
On 1.8 per cent of the party vote, Te Pati Maori’s proportional entitlement is two seats; holding six electorates lifts the party to six, and Parliament expands by the four-seat difference.
That is why the table totals 124 rather than 120, and why the majority line sits at 63.
In other words, the poll already contains one overhang — and it is the only one it counts.
Every other party is assumed to win no more electorates than its party vote would justify.
That single assumption is doing an enormous amount of work.
What the seat table assumes
Under MMP, the party vote decides how many seats a party gets; the electorate vote decides only who fills them.
For most of the MMP era that has made individual electorate races a sideshow to the national contest.
The overhang is the exception.
It is created not by the party vote but by the electorate vote — when a party wins more electorates than its share entitles it to. The surplus seats are bolted on top of 120, and the number needed for a majority rises with them.
The Verian table assumes National converts its 29 per cent into about 37 seats and wins no more electorates than that.
On the day, that is unlikely to hold.
I set out why in The overhang trap: at around 30 per cent of the party vote, National is on track to hold far more electorates than its proportional share, because the voters deserting it for ACT and New Zealand First on the party vote are still ticking the familiar National name on the electorate ballot.
The party vote goes one way; the electorate vote goes another.
The gap between them is a National overhang the poll does not model.
Scenario one: the overhang National is building
Take the poll’s own party-vote figures and add one modest, evidence-based assumption: that National holds four electorates more than its 37-seat entitlement — 41 seats won on the ground against 37 earned by the vote.
That is conservative against the six-to-eight-seat overhang I have argued is the likelier outcome, and it tracks the 2020 precedent, when National’s candidates held blue-rosette seats across Auckland’s northern fringe even as the party vote was being routed nationally.
The table below decomposes the same party vote into electorates won and list MPs returned.
Table 1: Seat projection — the electorate correction (National +4 overhang)
Party | Party vote | Proportional | Electorates | List MPs | Overhang | Total |
Labour | 32 | 41 | 23 | 18 | — | 41 |
National | 29 | 37 | 41 | 0 | +4 | 41 |
Greens | 13 | 17 | 1 | 16 | — | 17 |
New Zealand First | 11 | 15 | 0 | 15 | — | 15 |
ACT | 6 | 8 | 1 | 7 | — | 8 |
Te Pati Maori | 1.8 | 2 | 6 | 0 | +4 | 6 |
Parliament | — | 120 | 72 | 56 | +8 | 128 |
Proportional seats by Sainte-Lague on the poll’s party vote; total = the greater of proportional and electorates won. Left bloc (Labour, Greens, Te Pati Maori) 64 · right bloc (National, ACT) 49 · New Zealand First 15 · majority 65. Electorate and list splits are indicative; the proportional, overhang and total columns are the arithmetic.
National’s list empties to zero: at 37 seats’ worth of party vote and 41 electorates, there are no seats left over for the list.
Te Pati Maori’s six seats are all electorates, New Zealand First’s fifteen are all list.
Two overhangs now sit on top of 120 — Te Pati Maori’s four and National’s four — and Parliament expands to 128.
Why four extra seats change the government
Here is the counter-intuitive part.
National’s overhang does not help National govern.
National and ACT together hold just 49 seats — nowhere near a majority.
What the four extra National seats do is lift the bar everyone else has to clear.
A 124-seat house needs 63 to govern; a 128-seat house needs 65.
The left bloc’s 64 seats are a majority in the poll’s 124-seat house.
In a 128-seat house they fall exactly one short — and Winston Peters decides the government.
That is the whole mechanism.
The poll says the left can govern because it counts only Te Pati Maori’s overhang.
Add the overhang National is quietly building in its heartland and the same party-vote result produces a hung Parliament, with New Zealand First’s 15 seats holding the balance — left plus New Zealand First, 79; right plus New Zealand First, 64.
Nothing about Labour’s or the Greens’ vote has moved.
Four electorate seats, won by the losing major party, move the government.
Scenario two: if Opportunity crosses five per cent
The second thing the poll cannot see is sitting in its own results.
The Opportunity Party polled 4.6 per cent — four-tenths of a point below the threshold that would carry it into Parliament. A party that clears 5 per cent qualifies for proportional seats; at 4.6 per cent, with no electorate, its votes are wasted entirely.
The gap is comfortably inside the poll’s own margin of error.
Suppose Opportunity crosses.
On these numbers it takes six seats from the proportional pool — drawn one each from National, the Greens, New Zealand First and ACT, and two from Labour. National’s proportional entitlement falls to 36, so the same 41 electorates now produce a five-seat overhang. With Te Pati Maori’s four, Parliament expands to 129.
Table 2: Seat projection — Opportunity crosses 5 per cent
Party | Party vote | Proportional | Electorates | List MPs | Overhang | Total |
Labour | 32 | 39 | 23 | 16 | — | 39 |
National | 29 | 36 | 41 | 0 | +5 | 41 |
Greens | 13 | 16 | 1 | 15 | — | 16 |
New Zealand First | 11 | 14 | 0 | 14 | — | 14 |
ACT | 6 | 7 | 1 | 6 | — | 7 |
Opportunity | 5.0 | 6 | 0 | 6 | — | 6 |
Te Pati Maori | 1.8 | 2 | 6 | 0 | +4 | 6 |
Parliament | — | 120 | 72 | 57 | +9 | 129 |
Opportunity modelled at the 5 per cent threshold (4.6 per cent in the poll). Left bloc 61 · right bloc 48 · New Zealand First 14 · Opportunity 6 · majority 65. Left plus New Zealand First, 75; right plus New Zealand First, 62 — short.
The second kingmaker
The arithmetic now turns strange.
The left bloc falls to 61, the National-ACT bloc to 48. New Zealand First holds 14. And Opportunity holds six seats that belong to neither bloc — a party that has historically declined to declare a side.
Count the coalitions.
The left with New Zealand First reaches 75, comfortably past 65. But the right with New Zealand First reaches only 62 — short of a majority even with Peters at the table.
The incumbent coalition of National, ACT and New Zealand First cannot, on these numbers, reassemble itself. To govern, the right would need Opportunity as well.
Opportunity crossing the threshold does not just add a party. It creates a second kingmaker — and it is the right, not the left, that finds the door shut.
There is a sharper edge still.
A left bloc of 61 plus Opportunity’s six is 67 — a majority without New Zealand First at all.
Whether Opportunity would ever go there is a separate question; it has built its brand on refusing to pick a side. But the option exists, and its mere existence dilutes the monopoly on kingmaking that the overhang otherwise hands to Peters.
The four results, side by side
Set the four results out as a single row each and the pattern the poll cannot show becomes plain.
In the chart below the parties run left to right — the left bloc, then the swing parties, then the right bloc — with electorate MPs in the solid shade and list MPs in the lighter one.
The dotted line marks the majority, and it steps to the right as overhang enlarges the house.
Seats in the House across four results: 2023 election (123 seats, majority 62) · 1News Verian poll, June 2026 (124, majority 63) · Scenario 1, National overhang (128, majority 65) · Scenario 2, Opportunity at 5 per cent (129, majority 65). Solid = electorate MPs; lighter = list MPs; the triangle marks where the left bloc ends. 2023 splits per the Electoral Commission; later splits indicative.
Watch the left bloc’s edge — the triangle where Te Pati Maori’s seats end — against the dotted line. On the poll it sits just past the line: the left can govern.
In Scenario 1 the bloc is exactly the same size, 64 seats, but the line has moved to 65, and the edge now falls just short. Nothing about the left’s vote has changed; the majority has simply moved.
National’s solid block, meanwhile, grows while its list shading disappears — the overhang made visible — and in no row does any single party come close to governing alone.
The number nobody is measuring
Both scenarios turn on the electorate vote, and the electorate vote is precisely what almost no one is polling.
National party-vote surveys tell you very little about who wins a given seat, and electorate-level polling is scarce and expensive.
As I argued in Silly buggers, that blind spot is not neutral: in 2023 the only published Tamaki poll came from a single partisan-aligned firm, and the choice of what to publish — and when — becomes a weapon in itself.
It is into that fog that New Zealand First has parachuted marquee candidates — an All Black captain in Tukituki, a former Labour minister in Napier — into seats it has little hope of winning outright.
The point is not to win them. The point is what a strong third candidate does to the arithmetic around them: every National electorate that falls, to anyone, is one more brick out of the overhang and one more step toward the hung Parliament in which Peters decides.
Opportunity, meanwhile, is running its new leader in Mount Albert, a seat Labour held by eighteen votes in 2023.
A poll is a prediction with its most important variables left out. This one says the left can govern. Whether it actually can rests on four seats in National’s heartland and four-tenths of a point beneath Opportunity’s name
