Let me repeat that: Political polls are USELESS. Like sitting in a rocking chair, it gives you something to do but you don’t go anywhere.
All political polls should be published with a warning to viewers, readers or listeners that makes it clear the information about to be divulged is not accurate, not proven and only millimetres removed from guesswork.
Sensational headlines like: “Newshub-Reid Research poll shows Labour governing alone as National languishes in the 20s” should also carry a warning along the lines of: ‘The following contains foul language, offensive speculation, sex scenes and may encourage violence’”.
Recently, the pollsters have been shown to be completely wrong on three important events and not just a little bit wrong but wrong by a long way: Brexit, the 2016 US election and the last Australian election.
Remember polling is just questioning a sample of the population then extrapolating the answers using a model, not unlike the brilliant work of Professor Neil Ferguson OBE from Imperial College, London on COVID-19, who provided the figures that changed our world. His brilliant work on COVID is surpassed only by his equally accurate predictions on the 2001 Foot & Mouth outbreak (predicted 150,000 deaths – 200 actually died), 2003 Mad Cow Disease (predicted 50,000 could die – 177 did), 2005 Bird Flu (predicted 150 million could die – actual number world wide was 282), 2009 Swine Flu (predicted 65,000 British deaths. Swine flu killed 457).
That anybody in a position of power pays any attention whatsoever to anything this turkey says is bordering on criminal malpractice and the same applies to political polls.
In the interests of maintaining our rocking chair ride whilst going nowhere, I read Stephen Berry’s Insight: Politics article with interest. At least it’s an interesting and entertaining read with a well thought out and intelligent set of conclusions (or potential ones).
I thought I’d join in and do a broad analysis of the actual election results since 2001 and see if there are any numbers there that are of any use keeping in mind that to govern, a party needs 51% either alone or with a partner.
In 2017, Labour ended up with 36.9% (46 seats) and National 44.4% (56 Seats).
We all know what happened and we probably all share Judith Collins’s view that even if we believe in miracles, Winston isn’t going to be one this year.
Whether the Greens make it past the 5% threshold this year is anybody’s guess, but let’s just look at the actual major party percentages since Helen Clark’s historic win over Bill English in 2001:
2001
Labour (Clark) 41%
National (English) 21%
As a matter of interest, the smaller parties polled:
NZFirst 10%
ACT 7%
Green 7%
2005
Labour (Clark) 41%
National 39%
2008
National (Key) 45%
Labour 34%
2011
National (Key) 47%
Labour 27%
2014 (Key)
National (Key) 47%
Labour 25%
The highest single party percentage ever was Key’s 47% in 2011 and 2014 with Helen Clark’s 41% in 2001 and 2005 next best. It would be fair to say that National has scored higher percentages more consistently than Labour.
English, with 44.4% of the vote in 2017, should have been PM but wasn’t – and therein remains the catch. The smaller parties are critical to the outcome and there’s only one small party guaranteed to be there post-election day.
Labour will not get 51% + on election day. It’s never happened before and there’s no reason to expect it would this time. The vacant lot behind Ardern, Robertson and semi competent Hipkins will see to that. National won’t get 51% either even though they have more depth in the team.
The outcome hinges on two simple issues. Will either of the main parties lose or gain a significant percentage, and will the Greens be there after the election?
If the Greens are out, we will have a new government. If the Greens get in, Labour may be able to hold on.
I’m picking no Greens, a surprising result for ACT and a new National led government.
If you enjoyed this BFD article please share it.