Why dumping Luxon may be less risky than his party thinks

Friday, 6 March 2026

Luke Malpass is politics, business and economics editor

OPINION: After a hard week in a hard month, coming after a hard two years, the pressure is now really on Prime Minister Christopher Luxon.

Having bumped off a nascent challenge last year, Luxon’s support looked grumpy but stable heading into this year. That is no longer the case.

The latest Taxpayers’ Union Curia Research poll puts National at 28.4%. National Party sources say their internal numbers are higher — around a whopping 32% — but even that should be cold comfort.

The problem is that although the temperature came down after leadership rustlings, the party’s — and the Government’s — fundamental problems have remained: low polling support, political flat-footedness and a fair bit of foot-in-mouthness. And worst of all — from the National Party’s perspective — there is no evidence that Luxon can improve it.

Sitting underneath the political scuttlebutt is real world concern: People are struggling with bills, lack of opportunity, and an average economy. They perceive that the Luxon Government has not delivered on these.

Even worse, in private conversations National Party MPs never focus on Luxon’s strengths, his vision, or his case for leading the party and Government in the right direction. The arguments are always framed in the negative. There is no obvious replacement, the thinking goes. The transaction costs are too high. It isn’t clear someone else would do any better. The public will think we are focusing on ourselves, not them. Stability is best.

Much of the political wisdom within the Government comes from people who learned their craft during the Key years. The general view is that disunity is death and that only childish or desperate parties change leaders in office. Sometimes the case of Australia is invoked: that the six prime ministerial leadership changes in a decade had a destabilising and degrading effect on both parties.

But the fact is that the National Party is desperate. And if it is not, it ought to be. There is a pall hanging over the place that simply won’t fade. They are stuck in a rut and without the means to fix it. No-one in the caucus or the Cabinet seems to quite know where to start.

And while many backbenchers have their reservations about Luxon, most say they sense economic fortunes and business turning up a bit in their electorates. This adds to the confusion — while things seem to be getting better, National’s polling keeps getting worse.

The Key-era diagnosis is reasonable, but it is also mistaken. It is easy to be grown-up, take the high road and assume that, in the end, kitchen-table concerns will win the day if you focus on them — when you are polling in the mid-40s with a popular leader.

National is decidedly not in that position.

It is polling in the high 20s and its leader is historically unpopular — although some Nats point out that Jim Bolger was never popular. But that was 33 years and an electoral system ago.

No-one wants to move on the leader for all of these reasons, but they all know that something needs to give.

Luxon had a very poor week (“I don’t think it’s been a great week for the Prime Minister,” Nicola Willis said), where he somehow came across as both clueless and evasive on Iran. Couple that with the dreadful poll and the questions are back.

Questions such as: how well will he really go in an election campaign?

Again, within National Party circles the received wisdom is that Luxon actually likes people and enjoys meeting them, whereas Chris Hipkins is an awkward political geek who can barely make small talk. Look how poor Hipkins was during the last campaign, the argument goes.

I spent some time on the last campaign trail with Hipkins. He was doleful. On a trip to Dunedin the campaign was too scared to even take him onto the University of Otago campus because he might get heckled by students.

But this, again, is backward-looking. Hipkins inherited a campaign that had been conceived for Jacinda Ardern and not sufficiently changed for him. He will almost certainly be better this time around. In 2023 Hipkins was also defending a record — but not his. It was Jacinda Ardern’s.

Luxon, however, faces a far harder task. He can’t just be energetic, eat ice creams and wear pirate hats this time while talking about how he is going to fix things. He is no longer the coming man. Times are tough and he is now the one defending a record.

And that all assumes that by the time of the campaign the National Party vote will be high enough — or voters still tuned in enough — for the campaign to make a positive difference.

The most significant argument within National against getting rid of Luxon is the transaction costs of such a move.

That is: the public airing of ructions within the party, reduced standing with the public, internal fights spilling over and any bitterness that might linger.

At face value this argument carries significant weight. But these things are circumstantial and closer examination makes them look weaker.

Take the Australian example. After nearly 12 years of government and a strong run of prosperity, in late 2007 then prime minister John Howard was swept aside by Labor’s Kevin Rudd. In 2010, faced with free-falling popularity (as well as a good deal of internal loathing), Rudd was rolled by Julia Gillard.

She managed to hold on and return Labor to power in a hung parliament. After breaking some promises and amid a campaign of leaking and bastardry, Rudd ended up rolling Gillard again in 2013 — a move widely credited with saving Labor a large number of seats. That was because Rudd had remained popular with the public.

Then, in 2013, Tony Abbott was elected Prime Minister. He was rolled by Malcolm Turnbull in mid-2015 after Abbott’s polling began diving. Abbott had done some odd things — such as reintroducing knighthoods only to knight the Duke of Edinburgh.

Turnbull, then a popular and charismatic figure, received stratospheric poll ratings when he first took over and arguably should have called an election. After a series of missteps those ratings fell, but he went on to win the 2016 election.

In turn, he too was rolled in 2018. A challenge was launched by Peter Dutton, but Scott Morrison came up through the middle and became leader. He then delivered an election victory in 2019 that almost no-one in Australia thought was possible.

Changing leaders can and does work to win elections and improve results. Every Australian prime-ministerial change since Paul Keating rolled Bob Hawke in the early 1990s has basically been an electoral success. New Zealand voters are not that different from Australians.

The transaction costs also look uniquely low in Luxon’s case.

Usually they are high because leadership spills represent a fissure over the direction — or even ideology — of a political party. Or the public gets miffed because the person they elected gets tossed out. Or spills bring factional warfare out into the open.

Little of that seems to apply here.

Luxon does not have a faction and has precious few political friends. There are no obvious deep fissures within National or major factional battles. And he is very unpopular with the public.

He was the answer to a problem in 2021: National needed a new leader who would unite the party and not leave half its MPs worried about utu from Simon Bridges. He was very new to politics and the caucus coalesced around him for those reasons. As a politician he was basically an unknown quantity.

National rolled the dice and Luxon got into office.

There are always unknowns, but it does mean the transaction costs of rolling Luxon could be relatively low. There are no great ideological disputes within National — just poor performance from the leader, the seeds of which were sown in 2023 when he underestimated the task ahead and subsequently over-promised how he would fix New Zealand.

The political left likes to speculate about how X or Y could never happen because the liberals and conservatives within National wouldn’t agree to it. But that misunderstands the National Party in government — especially this one. There are the usual disagreements, but generally the party agrees on most things. It’s just a matter of degree.

Luxon’s problem is performance. He has not managed to gel with New Zealanders and, on consistent public polling, it looks likely to cost about a quarter of National MPs their jobs — and possibly the Government.

It might be bad management or bad luck. It is almost certainly unfair. But that’s politics.

So those are the reasons why, in Luxon’s case, rolling the leader might be the most sensible and least costly thing to do. The fact that both Labour and NZ First reportedly regard Luxon’s premature departure as their biggest risk speaks volumes. There are of course risks around coalition partners, and Winston Peters in particular.

However, there is one big reason not to roll the boss: it could always get worse.

Things will be economically tough — and may yet deteriorate further depending on the Iran war. Would a leadership change really settle matters? No-one has successfully rolled a sitting PM and been re-elected to tell the tale in modern New Zealand history.

It can always get worse. The question now is whether that is more or less likely with a change of leader.