It could absolutely happen – but a lot would have to go wrong first, explains Andrew Geddis.

As we move into the final week of the 2023 election campaign, there’s a near-unanimous consensus among those contesting it that the return of New Zealand First to parliament would be A Very Bad Thing. It’s not a total consensus, of course, because one Winston Peters and – if the polls are to be believed – some 5-6% of the population strongly disagree with the suggestion.

However, Labour, the Greens and Te Pāti Māori are very happy to repurpose the “coalition of chaos” line used against them ad infinitum and invite Christopher Luxon to try it on for size, as contrast to their own recent mutually admiring high-fives. National and Act can see their apparently secure march to ministerial office threatening to turn into a victory of the Pyrrhic variety, as the spectre of meshing their policy platforms with that of a fundamentally conservative nationalist party rises from the crypt. And the scattered minor protest parties must resent every percent of support that New Zealand First hauls in as eating the lunch they spent the Covid-years so busily preparing.

Yes, New Zealand First really is the dinner guest that no-one wanted to be invited to the table, much less get sat alongside. Except, and this really is very, very crucial, the 5-6% of the population telling pollsters they plan to support the party where it matters most – in the privacy of the polling place.

Which means, under our MMP electoral system rules, New Zealand First looks odds-on to be in the next parliament with anywhere from 6-9 MPs (depending on its final share of the party vote and the total size of the “wasted” vote). And, on current polling at least, there’s no clear avenue for anyone to gain the backing of a parliamentary majority without including those 6-9 MPs.

Which has National’s campaign chair, Chris Bishop, really ramping up the alarm bells about potential consequences. There is, he says, “a very real and growing possibility” that New Zealand First’s presence in parliament “would necessitate, essentially, a second election”.

While we might discount this claim a bit due to its political expediency – “Avoid chaos! Secure your future now!! Vote National (or, at least, Act)!!!” – there nevertheless is a foundation of possibility to it. How, then, might we end up having to revisit the polling places soon after already doing so?

Well, the first scenario Bishop points to is where the “left” bloc (Labour/Greens/Te Pāti Māori) and the “right” bloc (National/Act/New Zealand First) end up in a 60-60 tie. That scenario is the result of our (in hindsight, a bit silly) decision to have a parliament with an even number of seats. And it’s been discussed already, noting that the existence of our “party hopping” laws mean no individual MP could switch sides to create a majority for someone to govern.

If the 2023 election did produce this outcome, would Chris Hipkins and Winston Peters suddenly find that they could work together in government after all? Maybe they could – politics is, after all, the art of the possible. But if not, then literally no-one could form a governing majority and without this, another election is the only circuit breaker available.

Of course, the polls would have to move quite a bit before the end of this week to produce such a result. More likely, at the moment anyway, is a result that sees National and Act with more seats than the left bloc, but not enough to form a parliamentary majority. Meaning, any governing arrangements would have to include New Zealand First in some way, shape or form.

Bishop’s warning is that such arrangements might prove impossible to negotiate: “We are concerned there would be an inability to strike a deal in the interests of the country.” How realistic is this concern, in light of the nine successful post-election governing deals that we’ve seen in our MMP era?

Well, the heat and drama of the election campaign probably isn’t the best time to assess it. So let’s just remind ourselves of how that post-election government deal making takes place.

First of all, any governing arrangements are in the hands of the political parties themselves – there’s only very minimal formal rules about what they have to look like. And there’s a host of possible forms such arrangements can take: coalitions; enhanced agreements on confidence and supply; agreements on confidence and supply; co-operation agreements; etc, etc. Parties can be in government for some purposes, not in it for others, and so on.

The only non-negotiable is that when a vote of no confidence is called in the House, there must be more votes against that motion than are cast for it. And when the government goes asking for money from the House to run the country, there must be more votes in favour of giving that money than are against it. Whatever it takes to get this outcome is up to the various parties to work out.

Second, there isn’t a set time frame in which to make these arrangements. It’s likely that discussions won’t even start properly until after the special votes are counted and the official election result is released in early November. That’s because all the parties want to know exactly where the land lies in terms of their seats in parliament. And also, because everyone is pretty exhausted after the campaign.

And, third, the shape and content of the agreement between the parties is up to them to work out. It may be a minimal “we’ll vote confidence for now, but wait to see what next year’s budget will look like before carrying it on” agreement, as Act has mooted (perhaps even seriously). Or, it may be a quite lengthy and detailed coalition arrangement like New Zealand First has preferred in the past.

Whether the parties will come to any arrangement at all then becomes a matter of political calculus. On the one hand, what policy concessions and ministerial positions are going to be required to get each party to sign on? On the other hand, what are the risks of refusing to offer enough (or, refuse to take what is on offer) in terms of the alternative? Because if Bishop is right and a governing deal can’t be struck, then there’ll be no new government able to take over from the existing “caretaker” one that carries on post-election.

And, with no new government able to take over, at some point the party leaders will have to agree that Chris Hipkins (or whichever Labour MP is acting as caretaker PM) can tell the governor general that a new election is necessary. At which point – sometime early in 2024 – we get to do all this over again. A full rerun with candidates put forward, campaigns run, ballots printed, voting places established just like this year, all at the cost of somewhere north of $150 million.

With, of course, the added factor of the voters being able to decide who really is to blame for making them once again have to think about politics and voting accordingly. Because, that’s the ultimate pressure on the parties to find a way to work together – showing that you are unable to do so risks having the electorate judge you to be unfit to govern at all.