National leadership poll (sort of interesting but out of date)

A public poll on the National leadership is of limited value, because the leader is chosen by National’s 56 MPs only, and the poll was conducted before the leadership contest began. But it is a bit interesting, especially National supporter results.

The Spinoff Exclusive: Poll gives Judith Collins slim lead as preferred National leader

A UMR Research survey puts the polarising MP in the lead – but only slightly, and her favourability numbers are dismal, an area in which Amy Adams holds bragging rights.

The tussle to lead the biggest party in New Zealand’s parliament will be a tight one, if polling conducted largely prior to Bill English’s resignation and exclusively revealed to the Spinoff is a guide. Of the declared candidates, Judith Collins can boast the greatest support as preferred National Party leader, both among National voters and the wider public, though her lead over Steven Joyce is statistically negligible.

Not surprising to see so many ‘unsure’. The poll is split over eight MPs with a third ‘unsure’.

Notable that Mark Mitchell doesn’t feature, but that’s not surprising because the poll was almost entirely before Bill English announced he was stepping down, so before any candidates put their names forward.

Favourability ratings are also pertinent:

Collins is slightly behind Adams on favourability, but has twice the unfavourability with about half respondents seeing her unfavourably.

UMR Research, whose clients include the Labour Party, returned the results from its nationwide online omnibus survey, conducted between January 30 and February 14 (Bill English resigned on February 13). A nationally representative sample of 1,000 New Zealanders 18 years of age and over are surveyed. The margin of error for sample size of 750 for a 50% figure at the 95% confidence level is ± 3.1%.

The margin of error for National supporters will be much higher.

Labour jump in Colmar poll

Labour have jumped up to 48% in the latest 1 News/Colmar Brunton poll. National has slipped a bit with Greens and NZ First struggling.

  • Labour 48% (up 9)
  • National 43% (down 3)
  • Greens 5% (down 2)
  • NZ First 3% (down 2)
  • Maori Party 1%
  • TOP 1%

That will shake up the parties and pundits. It’s not a big drop for National but they will be worried about a downward direction with English stepping down, but it’s a major bump for Labour, at the expense of their partner parties in Government.

Ardern is up 4 to 41% in preferred Prime Minister, Bill English dropped but is irrelevant now in polls.

Newshub poll – National and Labour close, NZF slump

A typically misleading headline from Newshub on their first political poll of the year: Labour soars to popularity not seen for a decade

According to the latest Newshub-Reid Research poll, Labour’s polling is at its highest level since 2007…

That might be correct for a Newshub/Reid Research poll, if their polling methods hadn’t changed in ten years and they have allowed for margins of error. But other polls last year had Labour around the same level of support (43%, 43%, 45%).

However Labour has risen, at the expense of NZ First.

  • National 44.5% (up 0.1)
  • Labour 42.3% (up 5.4)
  • Greens 6.0% (down 0.3)
  • NZ First 3.8% (down 3.4)
  • TOP 1.4% (down 1)
  • Maori Party 0.8% (down 0.4)
  • ACT 0.2% (down 0.3)

Most of those results are too small to mean anything, except for the Labour lift and the NZ First slump.

Preferred Prime Minister:

  • Jacinda Ardern 38% (up 8.3)
  • Bill English 26% (down 9)
  • Winston Peters 5.7% (down 1.4)

Newshub-Reid Research interviews were conducted using online polling and computer-assisted telephone interviewing. Some 750 people were interviewed by phone and 250 online between Thursday, January 18 and Sunday, January 28.

Data is weighted to ensure a cross-section of representation of age, gender and geography.

The sample error is maximum of +/-3.1 percent at a 95 percent confidence level.

Also: Kiwis unfazed by Prime Minister’s pregnancy

And:  Bill English has solid backing as Opposition leader

One could wonder why Barry Soper raised the issue of English’s leadership in an ‘opinion’ piece this morning.

Hopelessly out of touch poll claim

Polls are often used to claim things that they don’t portray. There is no way of knowing exactly why polls move, and what timeframe cause and effect operates under. Pundits can only guess, or make things up.

In New Zealand media companies who publish polls try to make dramatic stories out of their own polls.

Here a niche blogger makes a ridiculous claim based on a US poll aggregator’s rolling results: Shithole countries comment gets Trump a big bounce in the polls

I had a chat amongst some other political tragics some weeks ago about how big a bounce would Donald Trump get with his shithole countries comments.

Some weeks ago? Trump made those comments just over two weeks ago, reported on 11 January (US date so 12 January NZ date).

It turns out a pretty substantial bump in the polls:

It doesn’t turn out to be anything of the sort. A Real Clear Politics ‘President Trump Job Approval’ chart is displayed – here is the same thing a day later, with the date of the shithole comment shown.

Since the comment there has been a small improvement in the poll average.  All polls cover several days and are obviously published after they are taken. Some of them are rolling polls. There is never a ‘before and event’ poll and an ‘after an event’ poll that can measure a movement on a specific day. So there is no way of knowing when a poll moves and why with any precision.

And different polls come out over time, with some leaning one way or the other, so the timing of the polls in the mix can make a difference, especially coming out of a time when some polls shut down for the holiday period.

RCP polling data shown at the blog:

I don’t know how that can tell anyone why a poll average changed due to one of many events that happened on 11 January, before that and after that. Trump is in the news a lot. A few days prior to his shithole comment Michael Wolff’s book was big news, and that’s likely to have some effect on poll trends.

A single rolling poll (Rasmussen Reports that tends to favour Trump) shows no appreciable change over the shithole period.

Date Approval Index Strongly Approve Strongly Disapprove Total Approve Total Disapprove
26-Jan-18 -12 30% 42% 44% 55%
25-Jan-18 -15 29% 44% 45% 54%
24-Jan-18 -16 29% 45% 44% 55%
23-Jan-18 -18 29% 47% 43% 57%
22-Jan-18 -18 28% 46% 42% 56%
19-Jan-18 -14 30% 44% 45% 54%
18-Jan-18 -16 28% 44% 45% 54%
17-Jan-18 -16 29% 45% 45% 54%
16-Jan-18 -16 29% 45% 45% 54%
15-Jan-18 -13 31% 44% 46% 52%
12-Jan-18 -13 30% 43% 46% 53%
11-Jan-18 -13 29% 42% 45% 53%
10-Jan-18 -14 30% 44% 44% 55%
09-Jan-18 -17 28% 45% 43% 56%
08-Jan-18 -16 30% 46% 42% 56%
05-Jan-18 -15 29% 44% 44% 54%
04-Jan-18 -12 31% 43% 45% 53%
03-Jan-18 -15 29% 44% 44% 54%

To understand what people thought of Trump’s comment requires a targeted poll. Like this one from HuffPost/YouGov:

January 11 – 12, 2018 – 1000 US Adults

According to a recent news report, President Trump asked “Why are we having all these people from shithole countries come here?” referring to immigration from African countries and Haiti. Do you agree or disagree with that comment?

  • Agree strongly 12%
  • Agree somewhat 14%
  • Total agree 26%
  • Disagree somewhat 13%
  • Disagree strongly 45%
  • Total disagree 58%
  • Not sure 16%

A comment at WO from ‘no bullswool’ would appear to be bullswool:

Donald Trump is refreshing in that he says what many ordinary people are thinking.

Back to the WO post:

Once again the media are shown to be hopelessly out of touch with ordinary voters.

Those are big changes over the previous months polling and you can clearly see his Approval ratings climbing rapidly off of the back of his shithole comments.

They are not big changes, ratings haven’t climbed rapidly, there is no way of linking minor poll fluctuations to one comment by Trump, and are a fool (or are trying to fool others) claiming you can see clearly what Slater is claiming.

Who is hopelessly out of touch?

US politics “has reached a dangerous low point”

A Washington Post/University of Maryland poll indicates that confidence in the US about their democracy is dropping, with many thinking politics has reached a dangerous low point.

Do you think problems in America’s politics right now are similar to most periods of partisan disagreement, or do you think problems have reached a dangerous low point?

  • Have reached a dangerous low point 71%
    (Temporary 31%, the ‘new normal’ 39%)
  • Similar to most periods of disagreement 29%

ShiningCityFlag

Do you think divisions today are at least as big as during the Vietnam War?

  • At least as big as during the Vietnam War 70%
    (ages 65 and over 77%, 18-29 65%)
  • Smaller 27%
    (ages 65 and over 18%, 18-29 34%)

How proud are you of the way democracy works in America?

  • 1996 (GSS) 16%
  • 2002 (Post) 9%
  • 2004 (GSS) 10%
  • 2014 (GSS) 18%
  • 2017 (Post/U-Md) 36%

How much, if at all, do you blame each of the following for causing dysfunction in the U.S. political system?

USPollDysfunction

Ordered by ‘A lot or some’/’Not at all’

Money is clearly seen as the biggest cause of dysfunction in US politics, but there are other contributing factors.

While Trump is well down that list he is the fourth highest ‘A lot’ at 51%.

Source: https://www.washingtonpost.com/graphics/2017/national/democracy-poll/

 

 

Dodgy ‘poll’ in Nelson

The Greens seem to be getting a bit dodgy in their attempt to get electorate votes in Nelson.

Patrick Gower: Desperate Greens drop fake news ‘poll’ in Nelson

The Greens have released the results of some phone canvassing which they’ve referred to as an “internal poll” that claims to show them ahead in Nelson.

But it is not a poll just like there wasn’t a fiscal hole.

It’s a set of numbers Green volunteers have gathered, with no way of checking them and media should be ashamed of reporting them as a “poll”.

It is not scientific, they have not released the raw data or methodology.

The first “question” is particularly dodgy: “Which candidate, between the Greens’ Matt Lawrey & Labour’s Rachel Boyack, do you think will beat Nick Smith?”

Now obviously this does not give voters the option of saying “Nick Smith will win” and is more of a push-poll.

It is actually “fake news” from the Greens.

It is a blatant bid to get publicity and get Labour voters to vote tactically and try and get Lawrey over the line to give them a lifeline in case they don’t make 5 per cent.

This certainly sounds dodgy, but there’s a certain amount of irony here.

Gower accusing someone of using a poll for “a blatant bid to get publicity” is a bit rich.

Poll: ‘The Big Issues’

In their recent poll Newshub/Reid Research respondents “were asked to rank the issues most important to them in this election, from seven major subject areas” – New Zealand’s top election concerns

This is a simplistic result – “the issue ranked most important to voters”:

That is a limited list, and gives no indication how much people are concerned about the issues. For example many people could be nearly as concerned about housing, the environment and money as about health. So this is a fairly rough indication of what people see as the big issues.

Here are party comparisons:

ReidResearchIssues2017Sept

For some reason they did not include a NZ First breakdown.

Who should NZ First go with?

The big question that Winston Peters won’t answer is how NZ First would go into coalition with, if given the opportunity to decide by voters and by National or Labour.

There is a chance they won’t get to decided, if they miss the threshold (they have dropped to 6% in the latest Reid Research poll), or if National or Labour can form majorities without needing NZ First.

Poll swings make it hard to know what the possibilities might be. Yesterday’s Reid Research poll has National very close to forming a government on their own, but the last Colmar Brunton had them short even with NZ First, and Labour could have got there with Greens and perhaps Maori.

Herald-ZB-Kantar TNZ has a poll of what voters generally think.

“If New Zealand First had the power to decide which party should lead the government after the election, on what basis should it decide?”

  • 38% – should go with the party which gives it the most policies, regardless of size
  • 35% – should go with the party with the most seats
  • 27% – not sure

Missing from the options:

  • Winston will please himself who he goes with

That’s the mostly likely outcome (in consultation with the rest of the party of course) so this poll is pretty pointless. Winston has a long history of dismissing polls that don’t suit him.

Newshub/Reid Research poll – huge reversal

The latest Newshub/Reid Research poll results show a huge reversal on the last Colmar Brunton polls.

Newshub: National could govern alone in latest Newshub poll

  • National 47.3% (last RR 43.3, last CB 39)
  • Labour 37.8% (last RR 39.4, last CB 43)
  • NZ First 6.0% (last RR 6.6, last CB 9)
  • Greens 4.9% (last RR 6.1, last CB 5)
  • TOP 1.6% (last RR 1.9, last CB 1.9)
  • Maori Party 1.1% (last RR 1.0, last CB 2.0)
  • ACT 0.6% (last RR 0.6, last CB )

djgnc1cvwaael-h

All the pundits will be busy revising all their scenarios.

But Patrick Gower is right about a key thing – party support appears to be volatile. Things could easily change over the next 11 days, although advance voting started yesterday.

Preferred Prime Minister:

  • Jacinda Ardern 31.7% (last RR 29.9, last CB 35)
  • Bill English  33.1% (last RR 30.1, last CB 31)
  • Winston Peters 6.9% (last RR 6.9, last CB 5)

The poll was conducted 6-11 September and has a margin of error of 3.1%.

The last Reid Research (RR) poll was conducted 22-30 August.
The last Colmar Brunton (CB) poll was conducted 2-6 September.

 

Chinese voter poll shows similar trends

A poll of 1300 Chinese voters shows strong support for national but also shows similar trends to general polls.

  • National 71.1% (down 2.4)
  • Labour 21.6% (up 5.8)

This is from NZ Herald with a silly headline: Poll: National will be back in Government if Chinese voters had their way

They don’t give any other party results but say that NZ First was down 2.4% and Act was down 2.0%.

The results are based on responses from 1300 Chinese New Zealanders who were eligible to vote in the September 23 election.

The WTV-Trace Research Chinese Poll is backed by local Asian media company World TV, and conducted by Trace Research Ltd, an independent market research consultancy.

Labour MP Raymond Huo has had a griizzle about the polling.

National List MP Jian Yang is believed to be the Chinese MP who would be the one to most effectively serve the Chinese community in the next three years on 44.8 per cent, followed by Labour’s Raymond Huo on 18.8 per cent.

Huo has written to the University of Auckland questioning the vadility of the poll, saying it “may not be robust enough to prevent it from some systemic abuse”.

“It appears to be nothing more than an online opinion survey ‘based primarily on the Chinese social media WeChat’ which is said to have more than 700 million subscribers worldwide,” he said in a letter to Professor Jenny Dixon, the university’s deputy vice-chancellor

Huo said the poll had been taken seriously because of Herald reports and its association with the University of Auckland, where Dr Zhu, is an honorary research fellow.

A university spokeswoman said the poll was carried out by an independent market research company and the university’s involvement was limited to a member of staff who helped proofread the Chinese-to-English translation of the results.

Zhu said Huo could “rest assured” that there were mechanisms in place to exclude those who are not based in New Zealand.

Zhu said it was common for elections to “bring out partisanship”.

Polls also.