Published in The Australian (Sydney), 13 October 2026
It is election time in New Zealand and as such, political tragics like myself cannot have any meetings without the conversation turning to the question “And, who will win on November 7?”
But something is different this year. Not just because the election is so tight with both the centre-left and the centre-right bloc neck and neck in the polls. The electoral system also adds a couple of uncertainties which make predicting the outcome a lottery.
You see, New Zealand operates under a rather unusual electoral system which combines proportional representation with having electorate seats. It is called MMP, which stands for Mixed Member Proportional, and it is just as straightforward as it sounds.
The basic idea behind MMP is simple enough: it is a system meant to deliver a parliament perfectly reflecting the will of the voters while also ensuring places around the country can choose who represents them.
The devil is in the detail – and this is what makes this year’s election genuinely unpredictable.
The first complication is relatively straightforward. Like in many other proportional representation systems, there is a threshold parties need to cross to make it into parliament. In New Zealand that threshold is 5 per cent of the vote. That means a party getting 5.0 per cent of the vote will have six MPs in a parliament of 120, while a party on 4.9 per cent will have none (unless they win an electorate seat).
Now, this year, there is a party that currently polls around that magic 5 per cent mark: Opportunity. It is widely grouped into the centre-left bloc of politics, alongside Labour, the Greens and Te Pāti Māori. That means that Opportunity’s fate matters to the centre-left outcome.
If Opportunity narrowly misses out, say on just under 5 per cent, the composition of parliament would be determined by the remaining votes for parties that made it. And that means that the centre-right bloc would receive two seats more than in a scenario where Opportunity just scraped in.
The problem is that in most recent opinion polls, the threshold sits within Opportunity’s margin of error. Which means it is really anyone’s guess whether they will make it or not.
Now, such threshold questions are common for small parties and easy to understand. But there is a second complication on top of that which is more puzzling: the possibility of an ‘overhang’.
That is different from the hangover one might feel after some elections, and it is a direct result of MMP. Or at least the way New Zealand designed its own version of MMP.
The problem is the interaction between seats won by local candidates and seats allocated to parties based on their proportional share of the vote. Under normal circumstances, and in every election so far, this caused only minor headaches. The two major parties usually won most of their seats directly and then got a top-up of list MPs based on their overall share of the vote. For the minor parties, except for rare occasions, it was usually list MPs only.
But this year could be very different because of an unusual combination of circumstances. Prime Minister Christopher Luxon’s National Party currently holds 44 electorate seats, but in opinion polls, it only hovers around 28 per cent.
Why is this a problem? Because 44 seats out of 120 total seats in parliament equals 37 per cent. But on 28 per cent of the vote, National should only be entitled to about 35 or 36 seats once the votes for parties that miss the threshold are set aside.
Of course, given their weak polling, National is likely to lose a few of the electorates it currently holds. How many precisely cannot be predicted, but pollster David Farrar’s seat-by-seat projections suggest there should still be around 37 or 38 seats left which National is likely to win. And that is still one to three seats more than the number of seats National should have on 28 per cent of the vote.
So, what all this points to is something that was not supposed to happen: an overhang for one of the major parties. If that happens, parliament simply grows, and National gets no list MPs at all.
Over the past weeks, I had conversations with four former National cabinet ministers. They were split down the middle when asked what they expect. Two said that the overhang would be a mathematical certainty, while the other two admitted the problem but still hoped for a miraculous lift in National’s polling that would somehow get one or two of their candidates into parliament via the list.
And these one or two candidates are not any candidates. Because of the way this election is playing out, they could be two of the most prominent frontbenchers National has: Finance Minister Nicola Willis and Housing Minister Chris Bishop.
Ms Willis does not even run in an electorate and needs her party to gain list seats to return to parliament. Mr Bishop, meanwhile, is one of the few National MPs who might lose his seat this election because his electorate of Hutt South has always been marginal for his party.
So it is complicated, and even more so because Opportunity’s result here will make a big difference to National’s prospects of bringing in candidates on its list.
The other parties have done the sums, too. Te Pāti Māori, polling near 1 per cent, wants only electorate votes while redirecting its supporters’ party votes to other parties of the left, so almost every seat it wins adds to the overhang. Meanwhile, both New Zealand First and ACT will hope that more National voters will vote tactically and increase their numbers of seats.
None of this breaks any rule. And that is the problem.
In 2012, the Electoral Commission recommended abolishing overhangs. But the idea was shelved because the parties could not agree on which changes to make. A 2023 review made the same case, again without success, and so the rule survives.
National’s official line remains “two ticks National”. Anything else, one of the former ministers told me, would confuse voters who after three decades still do not understand MMP.
Nobody can call this election, and on November 7, we shall find out whether New Zealanders understand MMP after all. But if they do, the proof may be a parliament that no longer matches how the country voted.
