Election 2023: All bets are off.

And why we need to improve our political literacy and numeracy.

The pundits have been talking as if the only post-election outcomes are either a Labour–Green–Māori Party coalition or a National–ACT coalition. Polls are currently suggesting that the latter could be the more likely of those two.

This preconceived binary view was reinforced, for instance, by Audrey Young in NZ Herald when she set out who’d be in a Luxon-led National–ACT cabinet – with exact allocations of ministerial portfolios. Perhaps Luxon is quietly grateful for her setting and completing his assignment for him. And, to be fair, she promises to do a similar service for Chris Hipkins. But Matthew Hooton pushed back on Young’s prognostication, arguing that ACT would demand more than the ‘morsels’ she’d given them.

It’s a free press, but I’d rate such pre-electoral speculation as not a good use of newspaper column space. And, as I’ll explain below, it’s unhelpful for the development of the MMP system.

On opinion polls last week we saw another example of disappointing punditry.

Newshub once again misled its audience with exaggerated click-bait about ‘blockbuster’ results. Unsupported by the evidence, they claimed on Wednesday that Labour was ‘on a downward spiral’ and that ‘Labour’s loss is National's gain’. Other media outlets parroted their messages.

But the lift of 1.3 points that Newshub cited for National since the previous Reid poll in early May was within the margin of error, and hence could be put down to random variation. It wasn’t worthy of comment other than: ‘no statistically significant change for National’. But that wouldn’t have got viewers.

While Labour has trended gradually (rather than spiraled) down from mid to low 30s since January, National remains stalled in the mid 30s. Last week, a Roy Morgan poll had National on 33.5 compared with Newshub–Reid’s ‘uppish’ 36.6%. Which of them was closer to the ‘actual’ level of voter intentions in favour of National? No one knows!

One reasonable conclusion from these polls, however, is that neither Labour nor National is in the 40s, where they’d like to be, and so this could be an election that favours the smaller parties overall.

Newshub could be accused of trying to shape voters’ expectations – and hence their voting behaviour – because they don’t objectively report survey results. They’re putting viewer numbers ahead of trustworthiness.

It would be in the public interest for the media companies who commission and broadcast political polls to declare how much they pay for them. This could help us understand the interpretation of polls as 6pm headlines.

---

Looking at some hard evidence, then, the combined party votes of National plus Labour in MMP elections haven’t yet fallen below 62%. In 2017, by contrast, the two major parties took 80% of the party vote between them.

In some comparable European proportional representation systems, the main centre-right and centre-left parties combined might get only around 50% of votes. And if a ‘winning’ party has only about 25% or less, then government-formation gets tricky and protracted. This shows us where things could head in New Zealand further down the track.

If last week’s Roy Morgan poll is anything to go by, with Labour on only 26%, we may need to anticipate (though not to predict!) Labour-plus-National capturing under 60% for the first time. That could presage European-style fragmentation, and a long-term decline of the major parties, accompanied by the growth of more radical parties.

But the Roy Morgan poll may have been an outlier, as the Reid–Newshub poll last week had Labour on 32.3% – more in line with other recent results, and certainly not ‘on a downward spiral’. The relatively low combined polling for the two major parties may just be a temporary effect in 2023, and 2026 could see a return to their joint domination of the electorate. Wait and see!

For now, the beneficiary seems to be the ACT Party which looks set to improve significantly on its 2020 election result of 7.6%. As their combined poll results are getting close to 50%, a National–ACT coalition government is a possibility. I just wouldn’t waste our time right now setting out their seating arrangements in a future cabinet! Such arrangements will depend on the parties’ numbers and relative bargaining power after the election, and on the concessions they’ll make over policies and office-holding.

There are possibly two more parties in the next parliamentary mix: NZ First and The Opportunities Party, both of which are knocking on the 5% threshold in some polls, or fancy their chances of winning an electorate (respectively Northland and Ilam). Voters disillusioned with the two major parties may go for alternatives, and a successful election for one or both would shift the overall balance of power.

Drafting up coalition agreements is too much like tea-leaf reading, so I’ll look at the post-electoral game as it’s been played in the past.

One of the problems that beset coalition negotiations in 2017 was the unwillingness of the Green Party to contemplate talks with National. That left the Greens captive to Labour and handed Winston Peters all the bargaining power. Not only did Peters ‘choose’ who governed, he was allowed to dictate the process of negotiations.

Now, the Greens had campaigned in 2017 to get National out of office, and their supporters would have held them to that. But it was politically naive of them to let Peters call the shots, become deputy prime minister and get seats in cabinet – and keep the Greens outside the tent with only a confidence-and-supply agreement.

The Greens could have undercut Peters’ bargaining power simply by offering to talk with National – without even making any commitments. Going by the numbers, National (56 seats) could have formed a coalition government with either NZ First (9) or the Greens (8).

If the Greens had taken the bull by the horns, they could’ve asked both National and Labour at least to consider greater policy concessions on climate and other environmental and social priorities.

Instead, Peters told everyone that they’d have to wait for two weeks after the election until final votes were counted. Meanwhile, he said he’d go fishing up home at Whananaki. Why didn’t the other parties start their own talks without him? Their minds were stuck in a one-dimensional 1970s model of the political universe.

If uncertainty similar to 2017 doesn’t reign after the 2023 election, then it may after the following election. Come such a time, it would be useful if the parties (and their supporters) adopt new approaches to post-electoral talks.

This could mean, for instance, not boxing the parties into government-formation commitments (‘who will you work with?’) before the votes have even been cast. Or, post-election, if it’s unclear who’ll form the government, anticipating that each party, without delay, should hold exploratory meetings with all others. And perhaps we could then drop the blather about kingmakers and skinny tails wagging fat dogs.

If the German example is any indication, we may one day have to accustom ourselves to a Labour–National coalition or to a range of multi-party deals. For instance, a sign of growing up would be TPM and ACT holding post-electoral negotiations. It’s a bit awkward that their tax policies are poles apart at the moment, but they have more in common on the Treaty than you might think.

It doesn’t help the country to adapt to MMP if opinion leaders are already ‘choosing’ cabinet ministers before the election has even begun. And it doesn’t help when they draw statistically unjustifiable conclusions from opinion polls. We deserve better.

---

If you like this, please share with others. There’s no paywall for the rest of 2023.