Predicting the election (badly), again.

19 September 2014

It’s become something of a tradition that, before every NZ General Election, I post my predictions to this blog.

Last time around I did OK. I didn’t think NZ First would get to 5% but my “B” predictions that included them were pretty accurate – only five seats out in total across all the parties.

This election is much harder to predict. It’s been the craziest NZ campaign I can remember and the polls have been all over the place. It’s hard to predict a result with any accuracy when the themes of the campaign have been so unexpected and unprecedented.

Polling in NZ is still in its infancy and there has been very little analysis on how they fare in a world where mobile phones are becoming more prevalent than landlines. There are also just four main polls in NZ, which means the potential for noise is higher than in countries with a larger population.

That said, I’ll give it a shot.

I think there’s about a 35% chance we have a change of government after the election. A lot of this is due to MMP and that Labour has far more options in forming a government than National. The latter will unquestionably be the largest party on Saturday night, but it could take weeks to unravel the results and for negotiations to conclude. I don’t think we will have a government until the end of September at the earliest.

National – The trend of the last three weeks has them dipping to the mid-40s and traditionally their support drops by a few percentage points on election night. It has been a terrible month for John Key and it’s not unreasonable to imagine him losing one-tenth of his base as a result of the corruption scandals.
My prediction: 43-45%

Labour – Personally, I think David Cunliffe has run a fine campaign but Labour remains low in the polls due to a variety of viable centre-left options. The party has focused on getting out the vote after the historically low turnout of 2011, and I think they will pick up more votes than expected from their early voting campaign.
My prediction: 25-27%

Greens – Polls have them (us, really – I campaign and vote for this party) around 13% but they have traditionally dropped on election night as their core support is young voters who are difficult to get to the polls. The campaign this year has been much stronger than previous years and I feel a result of 14% is achievable if everyone turns out to vote. Sadly, that probably won’t happen.
My prediction: 11.5-12.5%

NZ First – I goofed last election by underestimating the level of unstated support for Winston Peters and his result last time (6.59%) would have been even higher if people did not fear they were wasting their ballot. This election, there is no doubt that he will hit 5% and I can see an even better result given his supporter base is made up of people who do vote.
My prediction: 7-8%

Conservatives – Colin Craig just can’t really win. If he had been handed a electorate lifeline by John Key, the security of entering parliament would have meant the core “family values” demographic would have had more reason to vote for him knowing they were not wasting their vote. As it stands, he has yet to hit 5% in any poll but he’s close enough that people may consider voting for him.

This is a fascinating case study in Game Theory – essentially the closer a party is polling to 5%, the more exponential value a vote for them has as it’s less likely to be wasted. Voters are essentially gambling that others make the same decision as them - in this case to support a party hovering just under the 5% threshold on the understanding that if every Conservative sympathiser voted for them, they would top 5%. The danger for the party is that too few people take this gamble and give their votes to National instead.

I don’t feel comfortable predicting this. The same thing happened to NZ First last election and like Peters then, Craig is peaking at just the right time. I feel like enough soft right-wing National voters are prepared to give their vote to the Conservatives given the rewards (six extra right-wing MPs) outweigh the risks (wasted votes).

So, with a deep breath, my prediction: 4.3-5.2% and I honestly don’t know if they make it. If forced to choose, I would say they fall just short. Let’s say 4.8%

Internet Mana – Hone Harawira wins Te Tai Tokerau easily but It feels like the party support as a whole is rather transient and fickle. Had the election been held two weeks ago, when outrage over John Key had been at it’s peak, I would have picked a result above 2.5% but I just don’t think there’s the same level of committed support to drag them to that result. People will stay home.
My prediction: 1.6-2.1%

Maori Party – They should win Waiariki but will likely lose their two other seats in Te Tai Hauauru and Tamaki Makaurau to Labour. They latter is probably their best chance of getting a second MP, as there are simply too many stronger options for Maori voters in the party vote and their overall support is likely to drop significantly.
My prediction: 0.7-1.1%

ACT – Epsom voters have proven they know how to play MMP and I can’t see David Seymour losing here unless Labour/Greens voters really push back and vote for Goldmsith en masse. But the wider public has abandoned ACT. There’s abolsutely no incentive to vote for them when a vote for National does exactly the same thing with less risk.
My prediction – 0.2-0.6%

United Future – Copy and paste the ACT entry. Peter Dunne will probably win Ohariu (although it’s closer than Epsom) but there is absolutely zero chance the party will poll enough to bring in a second MP. It’s like they’ve given up even trying and probably the best result for the right will be if Dunne’s party polls so low that he becomes an overhang MP.
My prediction – 0.1-0.4%

So, putting all of these calculations into a calculator, this is what I come up with. This is my final prediction.

National – 57 seats
Labour – 34 seats
Greens – 15 seats
NZ First – 10 seats
Internet Mana – 2 seats
Maori Party – 1 seat
ACT – 1 seat
United Future – 1 seat (overhang)

A majority of 61 seats is needed to govern. National could not form a government without NZ First. Alternatively, Labour/Greens (with support of NZ First) would need the support of either Internet Mana or both United Future and the Maori Party.

Alternatively, let’s pretend that The Conservatives make it to 5.1%.

That would leave us with:

National – 55 seats
Labour – 32 seats
Greens – 15 seats
NZ First – 9 seats
Conservative – 6 seats
Internet Mana – 2 Seats
Maori Party – 1 seat
United Future – 1 Seat (overhang)
ACT – 1 Seat (overhang)

A majority of 62 seats is needed which leaves John Key being able to form a coalition with some configuration of the Conservatives, NZ First, Act, Maori Party and United Future.

The largest possible left-wing coalition (Labour, Greens, NZ First, Internet Mana) would be stick at 58 seats.

Neither scenario promises solid government. I’m convinced Winston would prefer to go into a government with Labour rather than National but if faced with the alternative of forcing a new election, he may be left with no choice.

Essentially, the election boils down to several key factors:

- Can National keep their support above 44%? Anything lower is incredibly dangerous. Time will tell how damaging the corruption scandal has been.

- Can Labour/Greens turn out their vote above 40%? If they can, they are likely to be able to govern. This is why both parties have focused almost all of their energy on inspiring non-voters to vote.

 Can the Conservatives top 5%? This would help National more than anything else.

- Will ACT and United Future be overhang seats?

Can the Maori Party win Te Tai Hauauru and Tamaki Makaurau? If they do, they are likely to be overhang seats and add to the centre-right tally.

- What is the real support level for Internet Mana. Has Kim Dotcom helped or harmed the party?

Any one of these can change the complexion of election night but the most likely outcome is that we will not have a winner on September 20. Negotiations will probably drag out for several weeks and the various configurations of parties depends on their demands and who can work with who.

Given how tight this all is, everyone should strongly consider voting. It's easy to do in NZ (just turn up to the polling booth!) and if you're overseas you have until 7pm NZ time to upload or fax your voting papers.

Readers of this blog will be aware I'm a Greens Party member and that I've spent the last three years campaigning for the Party vote amongst ex-pats here in the UK. I'm excited and nervous about this election but I really think there's a real chance to change the government and the biggest hurdle to that is voter apathy. In an election this close, it does matter. 

I've writttn at length before about my reasons for being part of the Greens and you can read my personal mission statement here.

I also recently wrote about why I campaign from overseas here.

And if you're curious about what I actually do, I've written a round up of our campaign hereIt's been a busy time and I'm looking forward to a break (my "Things to do after the election" list is very long - and includes updating this blog more regularly). But it's been a rewarding, challenging experience that I feel proud to have been part of.

I'll be watching the election live in a pub in North London with my team and 30 other Kiwis. I've allowed myself to daydream a little about seeing the government change, and tried to prepare myself for the alternative. It feels strange that all of this is finally at an end.

On a personal note, thank you to my friends reading this who have been so understanding when I can't make it to dinners, or parties or just hang out. It's hard enough working full-time in London and surviving, let alone putting in 20+ hours on the side each week, but no one has ever told me to give it up. I've promised myself I'm going to take some time to myself and that means I'll be around more.

But for now, please, make sure you vote. Vote for whoever you like but make sure your vote represents you and your vision for Aotearoa. It only takes ten minutes but it lasts a whole three years.