Frequently Asked Questions
How many votes does a party need to win?
Under the MMP system it isn’t usually necessary to get more than 50 per cent of the party vote to get a majority because there is usually a percentage of “wasted vote” – those parties that fail to win seats and don’t cross the 5 per cent threshold. For instance, in 2014 National got 47.04 per cent but won 60 out of the 121 seats in Parliament.
How does a party win seats and get into Parliament?
The only way a party can win a seat is by getting above 5 per cent of the party vote or winning an electorate.
If your party doesn't get above 5 per cent of the party vote or win an electorate seat, your party vote doesn't get transferred to your next choice. Your vote just doesn't count in the allocation of seats in Parliament.
The seats those party votes would have represented get divvied up among the parties who did win seats in Parliament.
Isn’t there a possibility the Māori Party will work with National?
Good question. The short answer is yes. The Māori Party have made it clear they are willing to work with either Labour or National. However, the Māori Party have pretty much ruled out being in government with NZ First, which is currently the only likely scenario in which National will be able to form government. They have also indicated that they would prefer to be in government with Labour, and that their preferred option would be a Labour-Green-Māori coalition.
Why have you not included Conservatives, United Future or other parties?
As we are a small team with limited resources, we have chosen to prioritise analysis of the parties polling over 1% or likely to win an electorate seat.
In 2014, United Future gleaned just 0.22 per cent of the party vote and only entered Parliament thanks to Peter Dunne winning the Ōhariu seat. When Dunne pulled out of the race for Ōhariu on August 21, United Future's chances of being in the next Parliament went with him.
The Conservatives won a solid 3.97 per cent of the party vote in 2014. But since then the party has struggled to register any significant support in the polls.
For that reason, we have not included them in our analysis.
How do you know which parties are more or less likely to win seats?
Currently, there are only three public political polls: Roy Morgan, Colmar Brunton, and Reid Research.
The most important thing to know about polls is that trends are what matter. A single poll should not be given too much weight without considering how it fits into trends over time. That’s why we have based our analysis on Stuff’s Poll of Polls and RNZ’s Poll of Polls to identify trends across the polls over time and take an average across different polls.
Based on trends over time across these polls of polls:
- National are trending towards 41-43%
- Labour are trending towards 40-42%
- NZ First is trending towards 7-9%
- Greens are trending towards 5-7%
We have marked all of these parties as likely to win seats in Parliament based on their share of the party vote. The Green Party is closest to the threshold, but has polled above 5% in most polls in the past month.
None of the other minor parties are polling above 5%, but some are more likely than others to win an electorate seat:
1. ACT is expected to win the Epsom electorate.
The party has been below 1% in most polls since 2014. Assuming they do win Epsom, ACT would likely need to get about 1.2-1.3% of the party vote in order to get a second MP. To get a third MP they would need close to 2%. Our assessment is that ACT is likely to win Epsom, and therefore to have at least one seat in Parliament.
2. The Māori Party is also very likely to win at least one electorate seat.
They are polling between 1-2%, which would point to them having up to three MP’s as long as they win at least one electorate seat. Our assessment is that the Māori Party has a good chance of winning at least one electorate, and therefore will have at least one seat in Parliament.
3. Mana leader Hone Harawira lost his Te Tai Tokerau seat in 2014.
He has since struck a deal with the Māori Party not to stand candidates against each other in the Māori seats – giving him a better chance of ousting Labour’s Kelvin Davis to win back Te Tai Tokerau. If everyone who voted for the Māori Party candidate in the previous election voted for the MANA, Hone Harawira would have won. Our assessment is that MANA has a less certain chance of winning an electorate seat, and therefore have marked them as ‘maybe’ getting seats in Parliament.
4. The Opportunities Party has consistently polled at 1-2% in polls.
Unless there is a radical jump in their party vote – or an even more surprising win in an electorate seat – TOP are unlikely to make it into Parliament. Our assessment is that TOP has a low chance of winning an electorate seat, and therefore have marked them as unlikely to get seats in Parliament this election.
A note on assessments vs predictions
These are the best assessments we could make, based on the evidence available. But they are not predictions. Anything could happen, and sometimes does.
A note on thEse polls:
The Stuff poll of polls is an average of the most recent poll from each public polling company – Roy Morgan, Colmar Brunton (published by TVNZ) and Reid Research (Newshub). It is updated each time a new poll is published.
Polls 0-35 days old carry a weighting of 1; 36 to 70 days old a weighting of 0.67; and 71 to 105 days old 0.33.
The method is based on four assumptions: (1) simple modelling approaches give better forecasts than complex ones, (2) non-sampling errors are reduced by combining polls with different methods in equal proportion, (3) the presence of random effects means that weighting by sample size is not required, (4) people’s views may change, so if a poll is not replaced it will gradually become less relevant.
The methodology was developed by Stuff with Professor Malcolm Wright of Massey University.
The RNZ poll of polls is an arithmetical average of the most recent major polls from: 1 News Colmar Brunton, Newshub Reid Research, Roy Morgan New Zealand plus the unpublished UMR Research (for the Labour Party) and, from September 2016 to July 2017 only, Curia (for the National Party). The polls are ordered according to the midpoint of their polling period.
How accurate have polls been in previous elections?
Pretty accurate. A look at the past three elections reveals that the polls have, collectively at least, done a good job at predicting the outcome. With a couple of exceptions, the Stuff Poll of Polls had the support for each party right to within 1 to 1.5 percentage points.
The biggest error in this group of elections came in 2011, when it over-estimated support for National by about 4 points (51.5 per cent versus 47.3 per cent) and under-estimated support for NZ First by 2.5 points (4 per cent versus 6.6 per cent).
Aren’t polls inaccurate because they rely on landlines?
This is a common concern, but it’s not supported by the evidence. For one thing, some polling companies include internet users and mobile phones. A quarter of the Newshub-Reid Research polls are based on internet users, and Roy Morgan call mobile phones. And in any case all major polling companies weight responses to ensure different demographics are represented fairly.