Today Non- Subscribers get a FREE taste of what they are missing out on.
Have a read of this Insight Politics article then decide whether or not you would like to subscribe to a Silver subscription or upgrade your existing Basic or Bronze level Subscription to Silver.
The Polls: The Sequel
Pity the poor ADHD political commentator with two debates and some exciting poll data all on the eve of a deadline. On second thoughts, don’t pity me because I get paid for this and can afford my own box of tissues.
In any case, the highlights of the US Presidential election were irritating on their own. I can’t sit for two hours, never mind two hours of that. Trump has grown on me over the last four years, and I appreciate the art of interrupting your opponent more than anyone. Hell, I’m one of the worst offenders when it comes to finishing other people’s sentences for them. However, on Wednesday I finally understood why people hate me doing it.
The NZ Leaders debate went until my bedtime. While I hear it was much more energetic than the previous, there’s only so many ways I can describe Jacinda’s wardrobe; fifty shades of rolled up sleeves and zero watts of ironing.
You don’t need to be on the spectrum (political or otherwise) to be dazzled by the latest polling activity and I think that has been magnified tenfold by the absolute drought in releases during Lockdown 2.0. The Mr Berry Mr Berry Poll of Polls takes the average of the five most recent polls to smooth out the spikes and oddities that can happen to any poll but there’s only so much that can be done with the post-dog’s breakfast being served up prior to the 21 September Colmar Brunton poll: Two Roy Morgans, two leaked UMRs and a Stuff-Massey University poll. It’s like going to McDonald’s and eating Oporto.
Accepting each poll as gospel looks ludicrous if you graph the raw data chronologically (with my enthusiasm for percentages and graphs it’s no wonder I married an Asian):
I haven’t excluded the non-parliamentary parties to trigger anyone (except Advance NZ supporters). It just so happens that only those five parties register in all ten previous polls. Stuff-Massey registered an 8.1% spike for Act in mid-July while Roy Morgan recorded 11.5% for the Greens one week before a leaked UMR placed them on 3.2%. As I discussed in a previous article on polls at the beginning of August, polls actually are a relevant measure of party support when combined into a poll of polls though not when examined by themselves, even if one were to look at a trend from individual pollsters e.g. the notoriously erratic Roy Morgan.
Perhaps death from embarrassment is the reason their usual first of the month poll has not been released after the Greens registered 11.5% in September.
To reinforce the point of my theory, these two graphs show where the parties registered on the last six Mr Berry Mr Berry Poll of Polls as each new poll was released and the average of the most recent five changed.
This is National and Labour
And Act, Greens, Maori, New Conservative, NZ First and TOP. These are the only parties that have managed to register a result above zero for the previous six Poll of Polls and I simply cannot abide line graphs that do not have continuous lines.
This is where all seventeen registered political parties now stand (so the 1.2% of voters who haven’t had their party mentioned yet can stop pouting):
Is there a point to all this?
Yes.
In August I also determined the average variance between party poll results for the key minor parties:
New Zealand First will exceed poll expectations by an average of 0.9%.
The Greens will get 1.4% less votes than the polls predict on average.
Act’s variance is statistically insignificant at 0.1% but they’ve also achieved less than 1.1% for the previous three elections.
This was my prediction of the election result before news broke of new COVID-19 community transmissions and the election date was delayed four weeks.
Labour: 46% (59 seats)
National: 37% (47 seats)
Act: 6% (8 seats)
Greens: 5% (6 seats)
New Zealand First: 2.9%
Maori: 1%
New Conservative: 1.2%
TOP: 0.4%
I’m sure nobody will begrudge me making another attempt to predict the result now, though the exponential surge in support for Act makes that difficult. In 2008 they under-polled by 0.7% achieving 3.5%. The trend for the Act party in the 2020 election stands apart from any pattern in any election for any party since United Future in 2002, so I’m going to have to resort to an educated stab in the dark.
Labour: 45.1% (59 seats)
National: 33.5% (43 seats)
Act: 8.1% (11 seats)
Greens: 5.4% (7 seats)
NZ First: 3.3%
Maori: 0.5%
New Conservative: 2.2%
Advance NZ: 0.6%
TOP: 1.3%
It would be historically unusual for Labour to fail to win a second term. A Prime Minister has not failed to win a second term without dying since 1960. However, the Greens are absolutely the key here and they remain perilously close to the 5% threshold. They have over-polled up to 2.2% previously and based on my calculations, that would bring them down to 4.6%. If that vote all went to Labour, Labour would govern alone on 63 with National on 46 and Act 11 seats. That would be a historical impossibility as no party has ever governed alone under MMP, even in Germany since it was introduced in 1948.
The centre-right block will not come close to beating the centre-left block if the Greens remain in Parliament. If the Greens fail to get into Parliament, then I’d put my money on the historically unusual defeat of a one-term government occurring over the possibility of a historic first: Labour governing alone. If you have left-wing friends (and I’m curious as to how you get them to remain friends with you), then convince them to vote Labour.
Did you enjoy reading that?
Subscribe to a Silver subscription or upgrade your existing Basic or Bronze level Subscription to Silver today.
