Danyl Mclauchlan posits a conspiracy theory:

there is a pretty great opportunity to use the Maori seats to rort the system: if you had two Maori Parties, one that ran electoral candidates in six out of the seven electorates and only canvased for electorate votes, and another that had a safe seat in the seventh electorate and only canvased for party list votes in the other regions then you could, conceivably, end up with a dozen MPs (albeit with some overhang due to your electorate imbalance) and hold the balance of power in perpetuity.

Danyl’s scheme is essentially what the Greens tentatively proposed before the last election: green voters in Māori electorates consider casting electorate votes for the māori party candidate, and māori party voters cast their party vote for the Greens. The proposal was rejected by the māori party, which at the time was (in my view) a tactical error but a wise strategic choice.

It was a tactical error because of the efficiency argument (a positive-sum alliance permitting the two parties to extract more parliamentary representation from the same base of electoral support). But without the benefit of hindsight it was a good strategic move because the māori party’s whole point was not to be shackled to the ‘Pākehā’ parties, and its long-term survival still likely depends on its establishing its own persistent powerbase; one which could stand a chance of surviving even the abolition of the Māori seats. The only way to do that is to grow the party vote by strengthening ‘brand’ support among its electorate. (Also the proposal was made quite late in the campaign, and the potential for voter confusion was high.)

Those factors which made the plan a strong strategic risk for the māori party in 2008 now no longer obtain, or at least not so strongly. It has largely sacrificed its independence from the Pākehā political mainstream anyway, and could potentially lose considerable support for that reason. It may hang onto electorates, but it is likely that ‘brand’ support is lower than before. So with the benefit of hingsight, they might as well have gone with the Greens last time. Nowadays, they have a radical wing-man in Hone, who will work with them and who shares, despite all the rhetorical ructions, considerable common ground with the party’s other members and its foundational kaupapa. They can work together; Hone working to attract the ‘brand’ support for an independent indigenous party, the existing māori party maintaining their electorate positions and continuing to work within the mainstream.

I’m not convinced this will happen — as the Richards say in comments to Danyl’s post, a lot of it is personality-bound — but it could work in principle.

There’s one other factor, though. Although ACT and National have been collaborating in this way since Rodney Hide was gifted Epsom, the moment the māori party and a new radical wing led by Hone start doing it, the headline will be “Maaries rorting democracy to establish an apartheid state”. Those who’ve been benefiting from this sort of positive-sum electoral coordination for years will be those most urgently banging that drum and waving those banners. That’s a powerful disincentive.

Edit to add: Of course, there’s also a referendum on MMP on the day of the election. A scheme such as this would be an outstanding means of undermining MMP’s popular support and endangering its future.

L