The problem with opinion polls

Written By: - Date published: 7:38 am, February 14th, 2020 - 77 comments
Categories: act, election 2020, greens, jacinda ardern, labour, national, nz first, Simon Bridges, uncategorized, winston peters - Tags:

Colmar Brunton released a poll result last night.

The results were quite different to the recent Reid Research poll and were remarkably stable to the last Colmar Brunton poll which was conducted last November.

The Reid Research Poll had National at 43.3% and Labour on 42.5%.  The Greens were on 5.6% and NZ First was in the danger zone at 3.6%.

The Colmar Brunton poll had National remaining steady on its November 2019 result at 46%, Labour was up 2 to 41% and the Greens were down 2 to 5%.  NZ First was struggling on 3%.

The different results allowed the related media entities to use remarkably different ways to describe what is happening.  Clearly both cannot be correct.

Jacinda’s preferred PM rating shot up a remarkable 6% to 42%.  Bridges inched up to 11%.  Over three quarters of his party’s supporters do not see him as being viable.  How embarrassing.

My personal view is that current polling does not properly represent the electorate as a whole.  It misses poorer people.  But unfortunately poorer people are less likely to vote. 

So the polling tends to approximate the eventual result but not through design. The inbuilt bias approximates the bias caused by turnout.  Which is why turnout and the ground game of each party will be so important this year.

77 comments on “The problem with opinion polls ”

  • Comments are now closed

  1. Mike 1
    14 February 2020 at 7:51 am

    Interesting suggestion but what leads you to the conclusion it is missing poor people? If you have something to back that up I would be interesting in looking as well. On a side note it is pretty unusual for a leader of the opposition to have a high prefered PM rating. Even Clark was just tredding water for a long time.

    • Sanctuary 1.1
      14 February 2020 at 8:36 am

      Because they always do – firstly, it was 50% landline. That will not only exclude most poor people but also automatically exclude practically anyone under 50.

      Poor people do not like interacting with anything they perceive as authority or information seeking. When you life consists of bailiffs, debt collectors, late bill payment calls and you are mostly on the run from debt you are unlikely to respond well to a random person asking you personal stuff.

      Now, a polling company will weight the poll. What they don't tell you is that if they only got three poor people when they needed nine they'll weight those poor peoples results by 300% to get the correct sample set. But obviously, that greatly increases the error possibility.

      Also, anyone know the undecided?

      • Gosman 1.1.1
        14 February 2020 at 9:24 am

        Have you evidence this is the case? If your theory was correct then parties which attract poorer people (Labour, Mana, and the Greens) would be getting consistently higher actual results than the Polls predict. I am pretty confident that is not the case and in fact the Greens tend to do worse than the Polls put them.

        • alwyn 1.1.1.1
          14 February 2020 at 10:24 am

          "parties which attract poorer people (…….. and the Greens)"

          If my experience is a fair reflection I would say that the people who support the Green Party are certainly not in the "poorer" category of society. The tend to be, usually as State employees, well on the upper side of the middle income bracket.

          They don't seem to understand that incomes above $100,000 per annum are not in the "poor me, I am so underpaid" category.

          Mind you, I live in the Wellington Central Electorate which used to have one of the highest average incomes, and one of the highest Green Party voting percentages in the country. Not quite as high perhaps as its neighbor Ohariu but still well above average I would think.

          • Gosman 1.1.1.1.1
            14 February 2020 at 10:41 am

            Point taken. Perhaps I should have stated "Parties that attempt to attract poorer people". 🙂

            • Wayne 1.1.1.1.1.1
              14 February 2020 at 4:17 pm

              Do the Greens even try to attract poor voters? I don't think so, not in any meaningful sense.

              Way too many of the Green Party issues are the concerns of the "woke" to be of any real interest to poorer New Zealanders, especially to the families of manual workers.

              I reckon the Greens know exactly who they are targeting. It is environmentally and socially concerned middle New Zealand who not immediately worried about their own economic situation. Especially those in secure state paid jobs. They will never suffer the economic disruption that green economic policies would bring about.

              For instance the thousands of people we see doing hard manual labor jobs on road building and road repairs. Not a group that Green party policies will appeal to, but who are a significant part of manual workers. Similarly all the retail workers in stores like the Warehouse and Briscoes. Green Party policies are a direct threat to their jobs and many of them will be very much aware of that.

              • You_fool …
                14 February 2020 at 4:31 pm

                I think you have a blue-green colour blindness issue

                • Macro …
                  14 February 2020 at 5:29 pm

                  Well Said Y_f .

                  Wayne has NFI who comprise the majority of Green Party members, or what their concerns are.

                  BTW Wayne I spent my last few years before full retirement, working on the roads – and it was one of the best jobs I ever hard.

                • Muttonbird …
                  14 February 2020 at 6:05 pm

                  +1.

                  The Greens clearly state environmental and societal issues are linked and of equal importance.

                  This sticks in the craw of National Party people because they'll not be able to get Greens support while this philosophy exists. The Nats need to split the Greens which is what they’ve been working on for some time both using the media and through paid proxies.

                  Wayne is being deliberately ignorant in this case because recent policy wins and policy development on social issues such as housing and tenancy security have come from the Greens.

                  Also the Greens have not stood on the way of infrastructure development in the way Wayne suggests. Quite the opposite in fact because they are part of a government which is doing a lot on infrastructure bringing both roading and rail in NZ into the 21st century.

                  He knows this but is dabbling in a bit of fake news I think.

              • Binders full of women …
                14 February 2020 at 8:16 pm

                True dat. GP get about 8,000 party votes in leafy Devenport and about 300 in Manukau. Poor people hate GP.

              • swordfish …
                14 February 2020 at 9:38 pm

                .
                Wayne … while I broadly agree with your implicit critique of the uber-Woke … evidence from the New Zealand Election Study suggests the Green support-base is not particularly affluent.

                See my detailed 2019 comment here https://thestandard.org.nz/can-the-greens-rise-like-the-liberal-democrats/#comment-1643383 …

              • Sacha …
                14 February 2020 at 10:31 pm

                They will never suffer the economic disruption that green economic policies would bring about.

                Just wait until they witness the disruption brought by not adjusting fast enough. We all know it will hit the wealthiest last (except maybe some coastal property).

        • Muttonbird 1.1.1.2
          14 February 2020 at 6:22 pm

          Clearly, the same conditions which see poorer people reluctant to participate in polling also apply when the time comes to vote.

          They are disenfranchised and will not seek to do anything which they perceive might draw attention to themselves.

          This explains Sanctuary's and others' points, that both polling and voting misses poorer people.

          Anyone with the slightest comprehensive skills would get this. Apparently not you, though.

      • peterh 1.1.2
        14 February 2020 at 10:05 am

        17%

      • Phil 1.1.3
        14 February 2020 at 10:13 am

        What they don't tell you is that if they only got three poor people when they needed nine they'll weight those poor peoples results by 300% to get the correct sample set.

        This is a factually incorrect description of the process. Polling companies don't just sample, say, 1000 people in the hope they get the right mix of voters and shrug if they only get 3 'poor' people when they need 9. There is a targeting process that goes on where they'll specifically look for groups of voters that are hard to reach and focus efforts on them to get closer to the turnout model they're using.

        • mpledger 1.1.3.1
          14 February 2020 at 7:05 pm

          I'd be very surprised if that is true. It would be too costly. From what I have experienced they mostly target by age and sex. The main problem is that they don't adjust for household size and they tend to get the "stay at homes". It doesn't seem to matter though because they get near the outcome at election time.

          • Phil 1.1.3.1.1
            17 February 2020 at 10:02 am

            Age, sex and ethnicity are all commonly used benchmarks. There's also the entire back-catalog of company data used to call back specific groups of people for second or third or more times to improve the pool of responses.

    • Gosman 1.2
      14 February 2020 at 9:22 am

      The old "But the Polls don't really reflect the general population" was put to rest over the last couple of elections. The results (other than a few outcomes, such as an underestimation of NZ First support and an overestimation for the Greens) has generally reflected the election results within a small percentage difference.

      • Nic the NZer 1.2.1
        14 February 2020 at 9:49 am

        Precisely none of the 2017 polls predicted the election outcome.

        • Phil 1.2.1.1
          14 February 2020 at 10:01 am

          The last polls for Reid and CB missed National's vote by a point and a half. They basically nailed Labour's vote. All the minor parties were well within normal parameters. Usual slight overstatement of the Green vote, which is a weirdly persistent phenomena.

          • Craig H 1.2.1.1.1
            14 February 2020 at 5:26 pm

            My guess is that undecided voters don't vote Green much, so the number of votes stays the same but the percentage share drops.

      • You_Fool 1.2.2
        14 February 2020 at 11:23 am

        Only thing I notice as being wrong is that ACT tends to be predicted to get 2 seats, but normally scrapes in with just the 1 seat. This is important because TVNZ made their prediction of a NAct government on their poll based on the 2 ACT seats.

        I notice also that the last 2 elections seem be filled with polls predicting a "close" election that is "going down to the wire" "too close to call" and all that, and that the various political "experts" need to on hand to tell us all what it all means….

        • Gosman 1.2.2.1
          14 February 2020 at 1:14 pm

          Incorrect. ACT for the past three elections has only been predicted to get 1 seat.

          • You_Fool 1.2.2.1.1
            14 February 2020 at 2:48 pm

            Maybe it is just me noticing the times they had predictions of 2 seats then…. in any case, the likelihood of them getting more than their 1 seat is minimal to non-existent… even after their gun and free speech antics…

      • Muttonbird 1.2.3
        14 February 2020 at 6:43 pm

        An overestimation for the Greens

        This is plainly false. I chucked a few numbers together around the CB poll which seems to be the poll of the day

        https://i.imgur.com/kzN6aKf.png

        https://i.imgur.com/haljR30.png

        https://i.imgur.com/2J7NIN5.png

        https://i.imgur.com/eXhnYTX.png

        What it shows is that CB couldn’t hit the side of a barn particularly with respect to the National Party. They are constantly over estimated except in 2017. In fact the most accurate of CBs election polling has been in 2017 which is clearly the most uncertain and volatile election in a long long time.

        It also shows that the Greens are not over estimated in polling. Traditionally they have been massively under estimated.

        That only point you might have is that CBs polling has improved since they widened their methods.

        • alwyn 1.2.3.1
          16 February 2020 at 5:54 pm

          I don't really have the time at this moment to go into a detailed argument but the comment is usually that the Green Party vote tends to be overestimated in the final, near election, polls. The poll figures 9 months or a year out from election can't be judged as accurate or inaccurate, because the question is typically something like "If an election was held today …". If an election doesn't happen pretty soon after the poll it is impossible to talk about whether it was accurate or not.

          By the way your table for 2008 says that the Green Party got 11.5%. They did not. They actually got 6.72%. Your conclusion for that year is the wrong way round, particularly if you look at the late polls.

          https://www.electionresults.govt.nz/electionresults_2008/partystatus.html

          • Muttonbird 1.2.3.1.1
            16 February 2020 at 6:10 pm

            My bad. The 2008 election result line should read:

            National 44.93% (-5.43 CB 12 month average)

            Labour 33.99% (-1.08 CB 12 month average)

            Greens 6.72% (+0.52 CB 12 month average)

            NZF 4.07% (+1.61 CB 12 month average)

            So you only want to consider Green polling on the eve of the election? M'kay then, let's do that and not speculate as Farrar has done on what Green polling in February means for the 2020 election.

            I look forward to you never referencing opinion polling again until the week prior to an election!

            • alwyn 1.2.3.1.1.1
              16 February 2020 at 10:38 pm

              No I did not say, and do not think that polling 9 months out from the election is not of interest.

              What I do say, and what I said just above here is that if you are going to try and decide whether polling numbers for the Green Party over, or under, estimate their support you can only do it for the polls that are taken just before the election. That is going to be the only time when you can compare them with reality, which is the actual election result.

              I don't think it matters when all is said and done. I think that both NZF and the Green Party are not going to be in the next Parliament. NZF will be gone because they are corrupt. The Green Party will be gone because they have turned out to be total wimps who have achieved nothing. They have also sat silent while NZF behave so badly and will not risk their comfy seats in the Limo's by calling out for Winston to be dumped.

              There will only be 3 parties in the next, and subsequent Parliaments. National, Labour and ACT. No- mates Labour will return to the Opposition benches in October and we will at least get a start made, after 3 wasted years, in getting our infrastructure repaired.

              • Muttonbird …
                16 February 2020 at 10:47 pm

                Delusional.

                Those tables show Green polling varying wildly in the month before each election. It is false to say they always do worse on election day. That's a meme created by Farrar which you and others who have trouble thinking for yourselves dutifully follow.

                Most Kiwis can now see it is the National Party which is the party of corruption. They’ve shown this over several years now since John Key entered Parliament.

                I think you are going to be very, very disappointed at election time.

                • alwyn …
                  17 February 2020 at 3:06 pm

                  Actually they didn't vary widely and they did usually give the Greens slightly higher numbers than they got in the election.

                  I give you the Colmar Brunton numbers for polls where the polling would have been within the last month before the election

                  2008. Two polls at 8.0% and 9.0%. Average therefore 8.5% and the election was 6.72%. Greens did worse than the polls by 1.8%.

                  2011. Four polls at 9%, 9%, 13% and 10%. Average of 9.75% and the election was 11.06%. Greens did better than the polls by 1.3%.

                  2014. Four polls at 12%, 11%, 14% and 12%. Average as 12.25 and the election was 10.7%. Greens did worse than the polls by 1.5%.

                  2017. Four polls at 5%, 5%, 7% and 8%. Average was 6.25% and the election 6.3%. Pretty well spot on. That as of course a truly wonderful performance by James Shaw who hauled them back from what looked a certain demise.

                  • Muttonbird …
                    17 February 2020 at 6:39 pm

                    Nice inconsistent cherry-picking of data. The truth is (and I know you struggle with the truth) polling of the Greens is volatile.

                    Polling companies don't seem to be able to get it right, although as I have suggested CB is doing better now they are including more methods.

                    By the way, CB over estimated National's vote in the all important n number of polls you think are important depending on the point you are trying to fabricate, by about 2 points in three of the last four elections.

                    • alwyn …
                      17 February 2020 at 10:55 pm

                      Last comment on this

                      I was not cherry picking, although perhaps you were.

                      I chose Colmar-Brunton because you did. It was their numbers and only theirs you put into your pretty little panels, remember?

                      I chose the polls that would have been done within a month of the election because you claimed that "Green polling varying wildly in the month before each election", remember?

                      I used all the elections from 2008 to 2017 because you chose them, remember?

                      So how did I pick the data I did? I used exactly the same method, and, after correcting on of your numbers, exactly the same data that you picked.

                      You are clearly admitting that you are a cherry picker if that is what choosing this data means. Now why do you find it so hard to admit that you stuffed up?

    • swordfish 1.3
      14 February 2020 at 11:28 am

      .
      Mike:

      it is pretty unusual for a leader of the opposition to have a high prefered PM rating. Even Clark was just tredding water for a long time.

      However, with the long-term decline in strong party affiliation / partisanship since the 1970s, leadership attributes (& related valence issues) have assumed ever greater importance.

      Key point: in the run-up to the last three Changes of Govt, Opposition Leaders were either outpolling or, at the very least, highly competitive with the Prime Minister in the Preferred PM ratings.

      TV1 Colmar Brunton

      (7 Months before Election)

      2020

      PM Ardern 42%

      Oppo Bridges 11%

      (PM leads by 31 points)

      2008

      PM Clark 31

      Oppo Key 36

      (Oppo Ldr leads by 5 points)

      1999

      PM Shipley 27

      Oppo Clark 21

      (PM leads by 6 points)

      2017, of course, was a little different (English easily outpolling Little 7 months out) BUT last minute Oppo Ldr switch … saw Ardern highly competitive with English immediately (First post-Leadership change Poll):

      PM English 30

      Oppo Ardern 30

      (PM-Oppo Ldr Tied)

      TV3 CM / Reid Research

      (7 Months before Election)

      2020

      PM Ardern 38.7

      Oppo Bridges 10.6

      (PM leads by 28.1 points)

      2008

      PM Clark 32

      Oppo Key 29

      (PM leads by 3 points)

      1999

      PM Shipley 22

      Oppo Clark 19

      (PM leads by 3 points)

      2017

      PM English 27.7

      Oppo Ardern 26.3

      (PM leads by 1.4 points)

      • Enough is Enough 1.3.1
        14 February 2020 at 11:34 am

        Interesting that Shipley was out polling Clark.

        This really does show that the preferred PM poll means fuck all

        • swordfish 1.3.1.1
          14 February 2020 at 11:50 am

          .
          Yeah, but only by a few points … Ardern's a whopping 28-31 points ahead of Bridges.

          To repeat:

          Key point: in the run-up to the last three Changes of Govt, Opposition Leaders were either outpolling or, at the very least, highly competitive with the Prime Minister in the Preferred PM ratings

          • Dennis Frank 1.3.1.1.1
            14 February 2020 at 1:09 pm

            Seems a very good key point to me. Opinion polls simulate the public mood at the time, and the public mood is always rather ephemeral. Too many things are likely to happen between now & the election to change that mood.

            The grinning strategy has worked for Simon, getting him up into double figures. But signs of intelligence would be required to make him competitive and they often seem to be absent. People do expect a political leader to at least know what's going on – but Simon struggles to keep up.

            Slippage in support for the Greens is likely to be due to their ongoing attempt to present as Labour lite. Labour itself is lite enough. We don't need the Greens to be that silly. We need them to provide a genuinely better option.

    • mickysavage 1.4
      14 February 2020 at 12:34 pm

      Turnout in the Maori and South Auckland areas is always lower than normal. This skewers the result towards the wealthy end of the electorate and the polling tends to reflect the eventual result although not always.

      • Gosman 1.4.1
        14 February 2020 at 1:16 pm

        I suggest this is merely your opinion. The only election in the recent past where this may have been the case that I am aware of is the 2005 election result.

  2. Nic the NZer 2
    14 February 2020 at 9:00 am

    Poll results already attempt to include turn out. Its one of the many ways various polls asking the same question differ.

    The big problem with polls is that calling a bunch of people and questioning them need not estimate the outcome of people going to a polling station and voting. The two are different activities and even an honest group of people may behave inconsistently in the different events. Also the people you call may not reflect the positions of the people who could vote. Usually as above this is estimated via a fudge factor guestimating the portion of electors you could have called. All of these estimates could be the product of a biased (e.g miss-leading) sample.

    Also of note, these errors are not factored into what is called the margin of error. What the margin of error describes is (given your poll was an unbiased sample) if you ran that poll across all the people you could poll, what is the likely error due to having asked only a sub-set.

    It doesn't however tell you the polls error compared to the election outcome. This will always be a higher margin again.

  3. Gosman 3
    14 February 2020 at 9:17 am

    Just curious if ANY opposition party leader outside Ardern and Key has scored more than 15 % in the preferred PM stakes over the past 15 years.

    • Enough is Enough 3.1
      14 February 2020 at 10:09 am

      Probably not

      And no other opposition leaders other than those two have gone onto win an election

      • Muttonbird 3.1.1
        14 February 2020 at 12:57 pm

        Exactly. I'm struggling to see what Gosman's point was. I suspect he is also.

        There have been no other leaders so it's clear you cannot lead a party to power while at sub 15%, Simon Bridges.

    • Sacha 3.2
      14 February 2020 at 10:31 am

      It is a rubbish statistic and I wish everyone would stop reporting it as if it were meaningful. Net approval is far more relevant.

    • swordfish 3.3
      14 February 2020 at 2:13 pm

      .
      Gosman

      Just curious if ANY opposition party leader outside Ardern and Key has scored more than 15 % in the Preferred PM stakes over the past 15 years.

      Focussing solely on the 2 TV Polls:

      Brash

      Brash almost always above 15% in both TV1 & TV3 Polls throughout his term as Oppo Leader (frequently in 20s)

      Goff

      Hit 15% just before the 2011 Election.

      Shearer

      Hit 15% in 3 consecutive TV1 Polls between Nov 2012 & April 2013, then fell back again (never that high in TV3)

      English

      TV1

      2017 Dec 28%

      2018 Feb 20%

      TV3

      2018 Jan 25.7%

  4. mosa 4
    14 February 2020 at 10:15 am

    Why is Labour not on a higher rating than National ?

    In all the time the previous government was in office they consistently led in all of the polling data.

    You could argue it was because Labour was divided and unstable due too its internal squabbles , or maybe it did not present a believable alternative too the status quo.

    With all of the star attraction of Jacinda and the yawing social deficit after nine years of neglect i would have expected Labour too be leading in these polls but it is not happening.

    Even as confidence in the direction of the economy has risen.

    45% want National too lead the government and have Mr Bridges who apparently no one has confidence in as the PM.

    It is unusual too have such a popular leader leading a government that is behind the opposition consistently in every poll since 2017.

    Maybe it is a hostile media which does have a detrimental effect with negative reporting on a daily basis.

    John Key who reputedly could walk on water and the dour Bill English never seemed too suffer being behind in these samples of public opinion on the most prefered party in parliament.

    When will the public get in behind the Labour party and reward it for its efforts ?

    • Enough is Enough 4.1
      14 February 2020 at 10:25 am

      "Maybe it is a hostile media which does have a detrimental effect with negative reporting on a daily basis."

      I don't think it is that.

      When has there been a positive media story abut Bridges…ever?

    • Sacha 4.2
      14 February 2020 at 10:33 am

      Why is Labour not on a higher rating than National ?

      It is not first past the post so individual parties do not matter. The coalition of parties around Labour has generally been larger than that around the Nats for many years.

      • mosa 4.2.1
        14 February 2020 at 11:50 am

        " It is not first past the post so individual parties do not matter. The coalition of parties around Labour has generally been larger than that around the Nats for many years "

        Yes but too govern you need 61 seats for a majority and Labour and its possible allies were never able too get a majority until Winston lent his support in 2017 after Labours amazing increase in party support.

      • alwyn 4.2.2
        16 February 2020 at 6:01 pm

        That can hardly be true Sacha. After all, if that statement was accurate in 2008, 2011 or 2014 we would have had Labour led Governments after all three elections wouldn't we?

    • Gosman 4.3
      14 February 2020 at 10:38 am

      To paraphrase you.

      "Wah wah wah! Why doesn't the media get in behind and make the people like us more?"

    • ianmac 4.4
      14 February 2020 at 11:42 am

      Mosa, "Why is Labour not on a higher rating than National ?"

      Labour would be similar to National but Greens take up some of the Leftish vote.

      National would be a lot lower if they split into their true factions. ie Serious Right Wing Conservatives V Moderates.

      • You_Fool 4.4.1
        14 February 2020 at 12:01 pm

        Add in some proportion of NZF as well for the "we like Labour, but don't like the Greens" faction (however small that might be) and add the rest of NZF support to Nat for the "We want old school conservatives (but like actual conservatives, not fanatical nutjob ones)" faction

        I think that still leads to a slight advantage to the leftish side of the ledger….

      • mosa 4.4.2
        14 February 2020 at 12:02 pm

        Hey ianmac thanks for that.

        Sure their are different factions with the " main parties " but the Alliance was too the left of Labour in 1999 but Labour still managed too be nearly nine points higher than National that year and still slightly ahead if you factor in ACTs support.

        And off course you had the " middle " with NZF and United with electorate seats keeping them in contention.

        https://www.parliament.nz/resource/en-NZ/00PLLawRP99111/4d5c2cf501956d02710301e0b3284ae8bc5758f7

    • Sacha 4.5
      14 February 2020 at 3:23 pm

      a government that is behind the opposition consistently in every poll since 2017

      Again, not FPP. Governing coalition behind opposition coalition in this latest poll but not even the recent one by a different company let alone all since 2017.

  5. ScottGN 5
    14 February 2020 at 10:33 am

    Neither poll appears to have survived the 24 hour news cycle.

    And I note that on TVNZ online it was only a few hours before the headline grabbing “Nats and ACT have the numbers etc” was replaced with “Poll shows election is going to be tight” more in tune with general analysis.

  6. Jimmy 6
    14 February 2020 at 12:18 pm

    Only one poll that matters and that is in September.

    • Incognito 6.1
      14 February 2020 at 12:28 pm

      Sure, but voters need guidance and confirmation from the opinion polls.

      • Jimmy 6.1.1
        14 February 2020 at 12:57 pm

        I guess the polls may spur some people in to action eg. NZF and Greens voters who are close to the 5%. Its just that lately polls (overseas) have been pretty inaccurate.

        • Incognito 6.1.1.1
          14 February 2020 at 1:02 pm

          Polls are good to create a feel & vibe. The internal polls are not made public unless they’re ‘leaked’. Election campaigns is when we experience peak manipulation and my cynicism sensor goes off the scale.

  7. Bruce 7
    14 February 2020 at 12:26 pm

    The polls are very important for my wife and I believe for many of her cohort as she likes to be sure she's voting for the winning team.

  8. observer 8
    14 February 2020 at 12:42 pm

    As Colmar Brunton states on its website, the full details are available 48 hours AFTER the headline numbers are reported on TVNZ.

    48 hours is an eternity in news. By then, the narrative has already been established, by the network who pay for the poll. So fact-checking is not possible, until it is already irrelevant.

    For example, last year the One News headline story was Christopher Luxon appearing in the "Preferred PM" ratings. They decided that was the Big News, and so made a lot of noise about this, at the top of their bulletin.

    Only later was it possible for the rest of us to see the results, which actually showed numerous other politicians with similar ratings around 1%, including Bennett, Seymour … and John Key! But "Key on 1%" was not a headline, Luxon was.

    In short, the polls are professional and useful, the reporting of the polls is not. They are two very different things.

    • Phil 8.1
      14 February 2020 at 2:47 pm

      In short, the polls are professional and useful, the reporting of the polls is not. They are two very different things.

      Shouting "Amen!" from the rooftop.

    • McFlock 8.2
      14 February 2020 at 3:52 pm

      polls months apart are useless AF.

      There's no reasonable trend, there's no way to identify what events are affecting which poll "changes", and there's not enough of a collection to identify rogue results.

      And public polling is worse than useless – it turns the election into a horse race, and it's a side-discussion about whether they report current support or simply influence the next results.

    • mosa 8.3
      14 February 2020 at 4:40 pm

      Totally agree Observer.

  9. ianmac 9
    14 February 2020 at 1:41 pm

    Agreed observer.

  10. Observer Tokoroa 10
    14 February 2020 at 2:53 pm

    The problem with opinion Polls

    Polls are awacky guess. For Instance, the whole Maori Leadership might give its vote to the People who are mostly attached to the Wealthy Citizenry.

    Or the Chinese people may flood their votes in support of the poor and ignore the wealthy Landlords and Business People. In the year 2013, approx 500, 000 industrious Asians belonged to New Zealand.

    Or they may do the Opposite.

    I seriously doubt that Politicians who play endless silly games with journalists and whom appear attached only to more and more personal wealth – will die at the Polls.

    For the reason, that the lower middle and struggling groups of people are not happy with the smug "I'am alright Jack stuff".

    In fact, they detest the slave making Wealthy parasites. And will tell the Wealthy in very clear language.

  11. Chris T 11
    14 February 2020 at 3:23 pm

    Never particularly understood the fixation with the leader polling unless the party polling starts to nose dive.

    • Sacha 11.1
      14 February 2020 at 3:25 pm

      So irrelevant in our political system.

      • Chris T 11.1.1
        14 February 2020 at 3:40 pm

        Indeed.

        Unless the party polling turns dismal and it is obviously because the leader is a bit of a shit communicator the leader poll is a tad something to look at, but who gives a shit

  12. Observer Tokoroa 12
    14 February 2020 at 3:40 pm

    The problem with opinion Polls

    Polls are a guess. For Instance, the whole Maori Leadership might give its vote to the People who are mostly attached to the Wealthy Citizenry.

    Or the Chinese people may flood their votes in support of the poor and ignore the wealthy Landlords and Business People. In the year 2013, approx 500, 000 industrious Asians belonged to New Zealand.

    Or they may do the Opposite.

    I seriously think that Politicians who play endless silly games with journalists and who appear attached only to more and more personal wealth – will die at the Polls.

    For the reason, that the lower middle and strugglng groups of people are not happy with the smug "I'm alright Jack ".

    In fact, they detest the slave-making Wealthy parasites. And will tell the Wealthy in very clear language.

  13. Peter 13
    14 February 2020 at 9:49 pm

    So there are little percentage points difference between parties and groupings of parties and there are margins of error and 16% floating.

    Getting all "this is going to happen or that is going to happen' is nonsense.