Looking Forward With Tactics or Strategies

JD

It seems some National supporters got the message on strategic split voting.

In 2023 National scored 1,085,851 party votes versus 1,192,300 electorate candidate votes. This nine per cent differential is a reasonable proxy for how many National supporters split their vote.

In 2026, there was a change to the ratio: 706,000 party votes versus 992,000 electorate candidate votes suggests as many as 29 per cent split voted this time.

Interestingly it was NZF who benefitted most with a party vote increase of 36 per cent or 63,000 votes. ACT only picked up 18,000 extra votes, seven per cent more than 2023. However these should be noted only as the preliminary results for 2026.

The real negative, however, is the 706,000 party votes for National as they gave it exactly zero extra MPs. The 42 National electorate candidates who won (preliminary result again, more on this later) will be in parliament but none of the ‘list only’ candidates will make it.

‘So what’ you might say? But recognise just how many extra MPs a split vote strategy would have bought us if these wasted party votes had been allocated to either NZF or ACT, our fellow travellers on the right-of- center (ROC).

If 700,000 party votes shifted to ACT/NZF, it would have given them an extra 36 seats on top of the 25 they have already claimed, lifting the right bloc seats to 103 in an ‘overhung parliament’ of 167 seats, and a significant majority in excess of 60 per cent, enabling it to pass critical legislation unhindered.

But that begs the question – if the ROCs start tactical voting like this, what happens if Labour supporters do the same?

In that case, given that the party vote between Labour and National was virtually the same, a similar incremental 36 seats is created for the Greens and TOP together.

Add in the five TPM seats, which look welded on in perpetuity, and you get a left bloc number of 85 seats in an overhung parliament of 188 seats. Still a 55 per cent majority for the ROCs.

The key reason for this differential is the number of electorate seats held by each side. National hold 42 to Labour’s 21, which could be said to be a real indicator of where the country’s political allegiance actually rests.

Ignoring party votes and race-based seats and arguments about who are disenfranchised and who disadvantaged vs who not, of the 66* elected seats where the population has a direct say in who gets to represent them (democracy, I think it’s called). 43 are ROCs, that’s 65 per cent. Need I say more.

*This ignores the Māori seats, which are race based and therefore outside of democracy. In these, the preliminary vote count for TPM electorate candidates is 67,728, which, when compared to the eligible Māori vote of 600,000+ further reinforces the ‘outside of democracy’ viewpoint. If only 10 per cent of your population votes for you, how can you claim to democratically represent them?

But I also want to look at the wasted votes by ACT/NZF supporters, those who voted for electorate candidates with no chance of winning – which in my view is just as egregious as National supporters wasting their party vote on National List candidates who similarly had no chance of winning.

On current counting there are an estimated 190,000 NZF/ACT electorate votes in this wasted vote category. In many electorates it doesn’t matter, as the National candidate is far enough ahead of the alternative to not need these votes, but in some electorates it’s critical.

In Mt Roskill, the National candidate is behind Labour by 262 votes with 763 ACT votes wasted on the ACT candidate. Why stand a spoiler candidate when every vote is critical? Doesn’t make sense.

In Hutt South, National trails Labour by 889 votes with the NZF candidate sitting on 813 wasted votes with 23 polling places still to be counted [at time of writing].

Count this up over all electorates and an AI analysis tells me that somewhere between three and seven seats might flip if these wasted votes went to National.

And so to the logical conclusion: National should pour all of its resources into finding and supporting good candidates for its electorate seats where, since MMP began, the ROCs have won an average of 35 seats per election compared to 29 for the left.

They should campaign solely on the message of ‘Vote for your National candidate – party vote for your National partners ACT/NZF.’

In return NZF/ACT should agree to stand no candidates in marginal National seats, or indeed no candidates at all, save perhaps David Seymour. But even that’s a waste, as he would get in on ACT’s list anyway, and his presence in Epsom denies a National candidate a seat which, under this split-vote system, would be incremental for the ROCs.

Even if the other side pursue the same strategy, we would end up with a parliament of around 190 MPs, which should please a lot of pundits on both sides of the aisle, like Sir Geoffrey Palmer and David Farrar to name but two, who have argued for this on the grounds that it would make parliament so much more effective in its scrutiny of legislation and expenditure – and perhaps provide more of a counterweight to the public service bureaucracy that currently runs the country.

Moreover, this parliament would be dominated by the ROCs, usually with around 105 seats of 190, which, I submit, is the real reason to think it’s a good idea.

Tactics like election bribes win battles. Strategies like perpetual split voting win wars. 

Notes: 

2026 MMP: Two 700,000-vote transfer scenarios

Illustrative calculations using preliminary party-vote and seat figures discussed previously. In each scenario 700,000 existing party votes are moved, split equally between two recipients; electorate winners are unchanged. Figures are estimates, not a verified Sainte-Laguë recount.

Scenario A – National transfers 700,000 party votes

350,000 votes to ACT; 350,000 to NZ First. National retains its 42 electorate MPs despite losing nearly all party votes.

Estimated additional ACT/NZF list MPs: ~36 combined. National loses no list MPs because it starts with none. Its 42 electorate seats create a substantial overhang.

Scenario B – Labour transfers 700,000 party votes

350,000 votes to Greens; 350,000 to TOP. Labour retains its 21 electorate MPs.

Estimated additional Green/TOP list MPs: ~36 combined. Labour loses all 15 list MPs, but retains 21 electorate MPs as an overhang.

Comparison

Important: These numerical seat changes are scenario estimates inherited from the earlier discussion, not audited official allocations. Exact totals, total parliament size and overhang depend on a full Sainte-Laguë calculation using every party’s valid votes, eligibility thresholds and electorate seats. These scenarios do not describe voter behaviour or election predictions.