Two political polls with similar results

Newshub released a Reid Research a poll on Sunday with ridiculous headlines and claims. 1 News released a Colmar Brunton poll last night with less dramatic but still over the top claims. Polls are just polls, especially this far from an election, but they try to get value from the expense of polling by making stories out of them that aren’t justified.

Last time the two polled the biggest talking point was how different their results were. The Reid Research poll was regarded as an outlier, being quite different to any other polls this term.

The most notable thing about the polls this time is that the results are very similar, taking into account margins of error of about 3% for the larger results, and the fact that Colmar results are rounded to the nearest whole number.

  • National: RR 43.9% (+6.5%), CB 47% (+2)
  • Labour: RR 41.6% (-9.2), CB 40% (-3)
  • Greens: RR 6.3% (+0.1), CB 7% (+1)
  • NZ First: RR 4.0% (+1.2), CB 4% (+1)
  • ACT: RR 1.4% (+0.6), CB 1% (-)
  • TOP: RR 1.1% (+1.0), CB 1% (-)
  • Maori Party: RR 0.7% (+0.2), CB 1% (-)

I don’;t think it’s surprising at this stage to see National a bit ahead of Labour, Labour has had a mixed month or two and is struggling to make major progress due to the restraint of coalition partner NZ First.

Green support looks at a safe level, but is well below what they were getting last term (about half).

NZ First are still polling below the threshold and will be in a battle to stay in Parliament.

Is is fairly normal these days there are a number of borderline governing scenarios with these numbers, with National+ACT and Labour+Greens thereabouts but not certainties.

A lot may depend on whether NZ First make the threshold or not next election. Both other times they have been in a coalition government they have lost support at the next election.

Trends from Opinion polling for the next New Zealand general election (Wikipedia):

That shows the last Reid Research anomaly well.

Preferred Prime Minister:

  • Jacinda Ardern: RR 38.4% (-10.6), CB 38% (-3)
  • Simon Bridges: RR 6.7% (+2.5), CB 9% (+3)
  • Judith Collins: 5.2% (-1.9), CB 5%
  • Winston Peters: CB 4%

Ardern a bit down, Bridges a bit up but still a big difference.

Newshub also did a poll on performance:

  • Ardern: performing well 62.4%, performing poorly 23.1%
  • Bridges: performing well 23.9%, performing poorly 52.7%

UPDATE: 1 News/Colmar Brunton have also started asking a similar question:

  •  Ardern handling her job as Prime Minister:  +33
    approve 62%
    disapprove 29%
    don’t know or refused 8%
  • Bridges’ handling his job as National Party leader: -22
    approve 29%
    disapprove 51%
    don’t know or refused 20%

Ardern performance is well above her party support, while Bridges is well below National support (about half).

  • Newshub-Reid Research Poll was conducted between 2-9 October 2019.
    1000 people were surveyed, 700 by telephone and 300 by internet panel
  • 1 News-Colmar Brunton poll conducted between 5-9 October
    1008 eligible voters were polled by landline (502) and mobile phone (506)

So both now rely on some polling by something other than landline, Reid Research 30% by internet panel and Colmar Brunton 50% by mobile phone.

1 News link here.

Newshub/Reid Search links here and here.

The Newshun headline says “Jacinda Ardern, Labour take massive tumble in new Newshub-Reid Research poll” but a more accurate description would have been “Newshub poll looks more likely following last rogue poll”. It wasn’t a massive tumble for Ardern, more like a large correction by Reid Research.

Leave a comment


  1. Gezza

     /  15th October 2019

    The Newshub headline says “Jacinda Ardern, Labour take massive tumble in new Newshub-Reid Research poll” but a more accurate description would have been “Newshub poll looks more likely following last rogue poll”. It wasn’t a massive tumble for Ardern, more like a large correction by Reid Research.

    😄😄😄 Very good !

  2. lurcher1948

     /  15th October 2019

    You get the poll results you pay for right Newshub

    • Duker

       /  15th October 2019

      One in every 20 polls conducted by some like Reid research or Colmar will produce the headline results which are just plain wrong. Thats statistics for you . The smaller parties jump about so much its no so noticeable.
      You can see it was TV3- RR turn last time looking at the dots on the graph. I seem to remember some time back when polls were more frequent one of the polling companies just canned the whole poll and never announced the result as it was

  3. The Newshub-Reid poll may be a large correction but the more interesting results for me are those for NZ First – struggling for political survival and still below the 5% threshold, yet Winston still holds his cabinet post and still holds the country to ransom over confidence and supply.

    • Duker

       /  15th October 2019

      Struggling below the 5% threshold? 1% in these polls means 10 people. and as they exclude the 18% who ‘dont know’ is likely to be 7 or 8 people.
      US polls call 30,000 numbers to get their 3000 persons. And they often will use public information of those who ‘voted last time’ to randomly call first , as they have a large portion who arent registered and a good 50% registered may not vote in ‘non presidential’ election years

      • Gezza

         /  15th October 2019

        Still, they’re a bit of fun. When the results are bad for a party or party leader the leader usually says they never pay any attention to the polls, or have various excuses about why the ressults are wrong or not relevant – the fallback being the only poll that counts is the election (which, of course is true, except they can & do influence elections, imo).

        But often, at the same time, if the latest poll shows a drop in popularity of their opponent party or leader, in the next breath, they claim the poll reflects their poor performance – in other words, it’s accurate for them.. Bridges did this last night.

        Our polls are too infrequent & the sample sizes too low for my liking.

        • Duker

           /  15th October 2019

          Nothing wrong with sample size, I get the impression you don’t understand stats. 2000 sample makes very little difference

          • Gezza

             /  15th October 2019

            It’s not that I don’t understand stats it’s that I don’t think they always replicate or substitute for reality. A 2000 sample might actually make a “very little difference” but a “very little difference” is all that it sometimes takes for a change of government to occur.

            • Duker

               /  15th October 2019

              Compared to other countries who have things like preferential voting, electorates or electoral colleges our system with it’s party vote % deciding the makeup of parliament and polling doing the same. It’s not so bad.
              The only problem is the fake news narratives they wind out of ho hum poll numbers.

          • Blazer

             /  15th October 2019

            dead right Duker ..look at how accurate the U.S polls were in predicting a Trump presidency and the U.K ones re Brexit.You make a very compelling argument about poll samples ..for ..idiots.

  4. TVNZ are also now doing leader approval polling:

    • Gezza

       /  15th October 2019

      Yes, I saw that on 1news at 6. Handy little addition to the set.

      • Kitty Catkin

         /  15th October 2019

        We haven’t heard how the sex assault investigation that Jacinda knew nothing about is going.

        The result could be very bad for her ‘caring’ image.

        The latest immigration stuff-up doesn’t look good, and nor does the union disenchantment.

  1. Two political polls with similar results — Your NZ – NZ Conservative Coalition

Leave a Reply to Gezza Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: