Poll of polls: election is National's to lose

Nine months out from the general election, National has a chance of getting enough votes to govern alone, recent opinion polls suggest.

Prime Minister John Key today announced the date as November 26, ending months of speculation about whether he would hold an early election ahead of the Rugby World Cup.

Although election years usually throw out a few banana skins, National looks to be in a strong position, with three major opinion polls showing its support at 50% or higher.

A Roy Morgan poll conducted in January found support for National at 55%, compared to only 29% for Labour, a 26 percentage point gap.

Colmar Brunton’s December poll also showed 55% support for National, with Labour getting only 33%.

A Herald DigiPoll was slightly better reading for Labour, which at 37.2% of decided voters was still 15.2% behind National on 52.4%.

Of the minor parties only the Greens look likely to reach the 5% threshold, while New Zealand First, which Mr Key has ruled out working with, is rating between 2.5% and 4.5%.

The Act Party continues to languish, with support of between 0.9% and 1.5% in the three polls.


Roy Morgan Poll January 2011

National: 55%

Labour: 29%

Greens: 7%

NZ First: 4.5%

Maori Party: 2.5%

Act: 1.5%


Herald DigiPoll December 2010

National: 52.4%

Labour: 37.2%

Greens: 5.3%

NZ First: 2.5%

Maori Party: 1.2%

Act: 0.9%
 

Colmar Brunton December 2010


National: 55%

Labour: 33%

Greens: 5%

NZ First: 3%

Maori Party: 2%

Act: 1%

This article is tagged with the following keywords. Find out more about My Tags

Post Comment

6 Comments & Questions

Commenter icon key: Subscriber Verified

These polls are useless: they only give results for decided voters. How many undecideds are there? Who are they leaning towards? Are those with a party vote preference actually going to vote? and are they registered to vote?
All of these factors are important in accurate polling.
The only polling company doing it this way so far is Horizon Research. They have the various colaitions neck and neck, with about 8% stull undecided, on sample sizes nearly twice the size of most of the published polls.
See the detail here:
http://www.horizonpoll.co.nz/page/89/potential-vote-detailed-results

When are the other pollsters and commentators going to reveal the results for all the respondents they contacted?
And tape properly into the intended vote?

Otherwise their moinor party results are hopelessly affected by the margin of error, and their overall result does nothing to really indicate the voting intentions of the registered, intending voters.

Reply
Share

These polls are useless: they only give results for decided voters. How many undecideds are there? Who are they leaning towards? Are those with a party vote preference actually going to vote? and are they registered to vote?
All of these factors are important in accurate polling.
The only polling company doing it this way so far is Horizon Research. They have the various colaitions neck and neck, with about 8% stull undecided, on sample sizes nearly twice the size of most of the published polls.
See the detail here:
http://www.horizonpoll.co.nz/page/89/potential-vote-detailed-results

When are the other pollsters and commentators going to reveal the results for all the respondents they contacted?
And tape properly into the intended vote?

Otherwise their moinor party results are hopelessly affected by the margin of error, and their overall result does nothing to really indicate the voting intentions of the registered, intending voters.

Reply
Share

Whatever the polls say, I like John Key's style. He has made a call and is prepared to back it. up with clear rationale.

Reply
Share

It is amazing how accurate the polls were at the last election - YEAH RIGHT.

Why nor publish how correct they have been for the last 20 years or more. Then it will be possible to work out the real margin error.

Reply
Share

John Key is a clever man - he knows how to punt on almost anything - as he ddi for many years on the foreign currencies in the financial markets.

But he has yet to come up with one economic development policy that will help transform this sinking country. Selling SOE's is a short term hole plugging exercise.

So John, you amy need to dosclose soem new robust policy and initiatives shortly so that voters have some confidence in your strategic vision - whatever that may be - we don't know yet.

Or you could be a one term PM and leave a legacy that you won't be proud of.

Reply
Share

Comparing different polls and their questioning tactics is not the real issue here - it is the very nature of the polling instrument,the 'random selection' of those answering the questions and their episodic nature that limits their accuracy and hence usefulness in predicting an election outcome.

Polls are inherently problematic by their very construction, as they ask relatively uninformed people how they would vote today, without factoring in any of the information that could affect their voting behaviour between now and the election (e.g. expectations of the international markets on the NZ dollar, the likely effects of policy announcements yet to me made). That is, they answer the question 'who would you vote for today', rather than the more important one of 'who do you think will win on November 26'. Of course, they use the very unsophisticated 'who would you vote for today' question, because the individuals in the random sample they select have no special knowledge about future events to share other than their current (relatively uninformed) current view based on historic information. A poll is in essence 'bringing forward the election' from November 26 to today, and then seeing what the result would be. So it is a very imperfect proxy for the really valuable question of what will actually happen on November 26.

By contrast, predictions markets allow people with 'inside information' (e.g. economists, policy analysts, political party insiders) to put some of their money on the line by using that information to stack the odds of winning a 'bet'. Consequently, their inside information 'comes out into the open' in the form of the current price for a claim on the winnings, leading to predictions markets being described as pooling the 'wisdom of the crowds' rather than relying on snap results of popularity contests on the day a poll is taken from a random sample of people who by the nature of their randomness, have no special information to share other than their current personal choice.Predictions markets have been shown to consistently outperform polls in their accuracy in regard to long-run projections. As they operate in real time, the also show the immediate response to new information that changes expectations. They are thus relatively more effective in identifying which issues are shaping the election oucome than polls, which, as they are conducted episodically can only ever give a 'snapshot' view.

The NZ predictions market iPredict was substantially better than the polls at predicting the outcomes of the last election. It currently shows the probability of a National-led government following the election at nearly 80%, with a National party vote of around 45%. https://www.ipredict.co.nz/app.php?do=browse&cat=164

Other interesting 'properties' currently being traded on iPredict show - the likelihood today of Hone Harawira winning Te Tai Tokerau as a Maori Party candidate is 66%, compared to the previous projection of 54%, showing how the predictions market takes immediate account of the information that the rift in the Maori Party is closing. Also, the prediction that Rodney Hide has a 75% probability of winning Epsom is of course strongly consistent with the strong likelihood of National leading the next government, despite its predicted party vote being below 50%. That there has been no change recently in the projected shares of the party vote of all major parties despite the discontent in Te Tai Tokerau suggests that ' those in the know' consider that the Maori Party's vote is relatively secure despite Hone Harawira's dissent and the threat of party disciplinary action being taken - that is, the 'Hone problem' is predominantly a local rather than a national issue, and is anticipated to have little effect on the eventual election outcome.

Reply
Share

Post New comment or question

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.

NZ Market Snapshot

Forex

Sym Price Change
USD 0.7733 -0.0013 -0.17%
AUD 0.9501 -0.0003 -0.03%
EUR 0.6313 -0.0023 -0.36%
GBP 0.4955 -0.0002 -0.04%
HKD 5.9975 -0.0098 -0.16%
JPY 92.7030 0.1190 0.13%

Commods

Commodity Price Change Time
Gold Index 1195.4 -2.890 2014-12-19T00:
Oil Brent 61.4 1.580 2014-12-19T00:
Oil Nymex 57.1 2.910 2014-12-19T00:
Silver Index 16.0 0.096 2014-12-19T00:

Indices

Symbol Open High Last %
NZX 50 5527.8 5553.7 5527.8 0.25%
NASDAQ 4759.0 4779.2 4765.4 0.20%
DAX 9827.3 9924.0 9787.0 0.81%
DJI 17812.2 17911.6 17804.8 0.49%
FTSE 6545.3 6620.9 6545.3 0.48%
HKSE 23264.0 23478.9 23116.6 1.26%
NI225 17685.5 17692.6 17621.4 0.08%