Polls everywhere

Written By: - Date published: 9:29 am, May 30th, 2011 - 87 comments
Categories: election 2011, polls - Tags: ,

Plenty of polls lately. Do we really need so many? Last night we got a 3 News / Reid poll:

National slips a little in poll, but still going strong

National has taken a hit in the polls, following its budget which made cuts to the popular KiwiSaver scheme.

Labour has been the major beneficiary, it’s received it’s largest boost ever in our 3 News Reid Research poll.

The numbers are at Curiablog:

National 53.0% (-4.5%)
Labour 32.8% (+5.7%)
Green 6.5% (-1.2%)
ACT 2.2% (+0.5%)
Maori 1.6% (-0.9%)
United Future 0.2%
Mana 0.5%
NZ First 2.4% (-0.4%)

The big jump for Labour is mostly correcting a previous rogue (yes they really do happen). Also last night, a One News / Colmar Brunton:

National loses support after Budget – poll

Support for National has dipped after the Budget this month but people are more confident about the economy, according to a new poll.

Tonight’s One News Colmar Brunton Poll put support for National at 52 per cent, down from 54 per cent last month, while Labour remained unchanged on 34 per cent.

The numbers for this one are similar to the 3 News, so I won’t list them. Obviously the direction of the movement is good for Labour. Equally obviously the size of the move isn’t good enough yet, unless it is the start of a systematic trend. We can only hope, as the paucity of ideas exposed by the budget sinks in, that a trend does gather force. I don’t think (as some commentators are claiming) that these results mean that National have survived the sub-zero budget at all. These things take time to register with the electorate, and there is time enough yet.

While still on the subject of polls, notice that no one here at The Standard thought it was worthwhile running this nonsense last week:

Labour and National coalitions neck-and-neck

A new political opinion poll has the potential post-election coalition blocs neck-and-neck.

Horizon Research says its latest survey shows the National-ACT-Maori Party-United Future group with 42.7 percent support, and Labour-Green Party-New Zealand First on 43 percent.

Ahh if only! As we’ve frequently commented here, these Horizon surveys use such an unrepresentative methodology, and generate results that are so different from every other poll, that I don’t see how any supposedly reputable news organisation (hello 3 News?) can go about quoting them at face value. They may be marginally useful for spotting trends within their own data, but the raw numbers themselves are nonsense. Interestingly however, on Saturday, Horizon took some methodological potshots of their own:

Political poll results may exclude 30%?

…However, Horizon Research says an investigation may show some polling firms are excluding up to 30 out of every 100 people who respond to them from published results, providing a distorted picture of party support levels in the electorate as a whole. This could over represent support for larger parties and under represent support for minor parties.

Horizon says if 30% of respondents have been excluded from published results for the latest post-Budget DigiPoll this would be an example. In that event, it would be rating National at 16.3% to 19.6% higher than its actual support. Labour support would also be over stated, while minor parties’ support could be understated by up to about 3%.

Horizon says the differences arise because some of the polls are publishing results only for those who say they have decided which party they will vote for. This may include people not eligible to vote, not registered to vote and not intending to vote.

However, respondents who do not know, choose not to vote, and prefer not to say may be being excluded. This precludes providing a total picture of the electorate.

In its post-Budget poll of a representative national population sample of 2,254 Horizon found 15.6% undecided, 6.3% choosing not to vote, 3.9% wouldn’t say and 4.2% voting for minor parties (other than NZ First, United Future, Mana and ACT), a total of 30%.

I’m unconvinced. While individual polls can be all over the place, historically rolling average “poll of polls” results have been very good predictors. I don’t see any way that current polls can be out by as much as Horizon claims.

But on the other hand, I don’t think current poll results are set in concrete either. The election campaign hasn’t properly begun yet, National have no constructive ideas and are running an unpopular line on privatisation. There is plenty of time for things to change. Leftie commentators who have been getting a bit hysterical lately, like Matt McCarten and Chris Trotter, should take a few deep breaths, and have a little faith. Look out for Labour’s policy in the run up to the election. There’s plenty of work to do. Onwards!

87 comments on “Polls everywhere ”

  • Comments are now closed

  1. David Farrar 1
    30 May 2011 at 11:15 am

    I share your scepticism over the Horizon online panel methodology. Their results are so far away from everyone else.

    As for their criticism, about not including undecideds, they have a point, but also over-state the problem. If you see the full poll results of the public polls, they do reveal the level of undecided voters. Generally it is under 10% – nothing like 30%.

    Publishers use the percentage of decided voters, as that is comparable to an election result. And that is important. But the level of undecided voters is also important. One could have two polls which both show National at say 47.4%, Labour 42.1% and Others 10.5%. Looks like the same result.

    But if one poll had 5% undecided and one poll had 15% undecided then the raw results are:

    National 45.0% v 40.3%
    Labour 40.0% v 35.8%
    Others 10.0% v 9.0%
    Undecided 5% v 15%

    In the first poll you would conclude National likely to win as they already have 45% support locked in. In the second poll National only has 40% locked in, so victory not so likely.

    So Horizon are right that the level of undecided voters are important. But all the other companies publish their level of undecided voters so it is trivial to reverse engineer the raw results. And when you do, Horizon are still way of out kilter with everyone else.

    • Gosman 1.1
      30 May 2011 at 11:44 am

      David you are an ‘evil’ right wing blogger who only polls to provide your masters the information they want to hear. This reasonable discussion of the ins and outs of polling is obviously a smokescreen to confuse the good readers of The Standard about your true motives,

    • Lanthanide 1.2
      30 May 2011 at 11:46 am

      One commenter on here has already admitted to having 5 fake identities that they use to answer the Horizon “polls” (more like questionnaires).

    • todd 1.3
      30 May 2011 at 1:22 pm

      I’m unsure why somebody with obvious political allegiances to National should have anything to do with the polling process? Just like his master’s, conflicts of interest are apparently acceptable to Farrar!

      • Gosman 1.3.1
        30 May 2011 at 2:08 pm

        You do realise that there is also a polling company that has strong ties to the Labour party don’t you Todd?

      • Deadly_NZ 1.3.2
        30 May 2011 at 3:00 pm

        I do their polls but only once because to have to answer the questions more than once??? forget it.
        and I get 1 shot at a prize. But I do answer honestly to the best of my beliefs and prejudices, political and personal.

    • onsos 1.4
      30 May 2011 at 8:01 pm

      Your polls are as useless as Horizon’s. This is not to venerate Horizon! They are useless. Until you can do the job properly, don’t bother.

  2. bomber 2
    30 May 2011 at 11:46 am

    David and The Standard agreeing with each other? This is a sign straight out of the book of revelations – how quick we are to forget that these polls are not all seeing predictions and at times manipulate opinion rather than represent it. These polls only use landlines, not cell phones and in a recession where many are ducking and dodging debt collectors, these chaep braifart polls David promotes (and sells to the Government, so it is in his interests to promote them over the conclusions of Horizon) are missing a large chunk of the voting population. The many issues involved in Weighting are never explored either.

    If anything should prove that these polls in a recession are providing false returns is the Herald Digi Poll on the eve of Auckland Super city voting claiming that Len Brown and John Banks were neck and neck – how did that election pan out folks? It was a landslide to Len. It wasn’t anywhere near neck and neck was it?

    The NZ Herald-Digi poll can kiss my arse – http://tiny.cc/0rgi3

    These polls are not reflecting public opnion, they are manipulating them, and for Farrar who was recently pinged in a conspiracy with Whaleoil and Lusk to topple MMP – I have little faith in his proclamations of his position defending these Polls as an honest broker.

    The less said about the NZ Herald aoppointing a propagandist like him as their on-line political columnist the better. With media bias that extreme, it’s a surprise that Labour are rating above 20% support in these equally biased polls.

    • Lanthanide 2.1
      30 May 2011 at 11:50 am

      Here’s Farrar’s latest blog he somehow finagled getting with fairfax: http://www.stuff.co.nz/national/blogs/david-farrar-by-the-numbers/5072622/Sixty-five-seats-is-not-bulletproof

    • Gosman 2.2
      30 May 2011 at 12:03 pm

      Ah Mr Bradbury since you don’t have editing rights here perhaps you would care to deal with the issue that I have put before you on Tumeke.

      I ran in to David Farrar last week and put to him your allegation that he polls over the weekend so that he can get the results to the National Party caucus on Tuesday. He has categorically denied this. Do you have any evidence to the contrary?

    • r0b 2.3
      30 May 2011 at 12:34 pm

      David and The Standard agreeing with each other?

      Actually, it’s only David and r0b. Other authors here will have their own opinions.

      I agree with you bomber that the polls are flawed in who they sample. I agree that they probably under-sample left wing voters – but not by a huge amount. I agree that they become self-reinforcing, and are used to shape opinion almost as much as they reflect it.

      But I don’t believe that there is a systematic attempt by polling companies to distort polling results.

      One day we’ll get an election where the results are way out of line with (average) polling, and we’ll know that polling methodology is truly broken. Until that time, I’ll take them with a grain of salt, but I won’t ignore their fundamental message. The Nats are still way way way too popular. We on the Left can’t delude ourselves that this isn’t so, we just have to work harder to fix it.

      • lprent 2.3.1
        30 May 2011 at 12:51 pm

        My general view is that you can rely on them for trends, but that they are quite unreliable for absolute numbers.

        The landline issue is quite noticeable in Auckland that a higher deprivation index has a strong correlation with the probability of a household having a listed phone line (which all polls apart from the Horizon one rely on).

        What the polling companies are implicitly relying on with political polls is the correlation having a higher deprivation index also increases the probability of people not normally voting. Of course that tends to fall over in periods like 2005 where the prospect of Don ‘Extremist’ Brash running the country helps to encourage people out to vote who’d normally ignore it.

        However often their biggest inherent flaw is the way that they are mis-reported. For instance in a way that doesn’t show the undecided and spun to highlight newsworthy stories.

        • onsos 2.3.1.1
          30 May 2011 at 8:14 pm

          I don’t believe even that. The trends are difficult to trace, the confounding factors are prolific, and they are proliferating.
          Don’t get me wrong–I don’t think Labour are going to win the election–but political polling in this country is moribund. I wouldn’t believe any of them.
          The only place I’ve seen polls become useful is at http://fivethirtyeight.blogs.nytimes.com/. There it is not th polls that are useful, but the analysis of them. Our polling companies don’t release enough data to make this possible. I suspect this is because they would be embarrassed by the weaknesses revealed.

          • lprent 2.3.1.1.1
            30 May 2011 at 10:01 pm

            Our polling companies don’t release enough data to make this possible. I suspect this is because they would be embarrassed by the weaknesses revealed.

            I’m sure of that

            • Colonial Viper 2.3.1.1.1.1
              30 May 2011 at 10:02 pm

              Oh for someone to do a wikileaks on the polling methodology…

              • Draco T Bastard …
                31 May 2011 at 2:13 am

                Well, here’s Horizons. Of course it doesn’t address the fact that it’s possible to register several times to be on their poll.

      • onsos 2.3.2
        30 May 2011 at 8:08 pm

        There is no deliberate attempt to weight results for political purposes–except perhaps by DPF, who is a tool–but the results are systematically biased regardless. It comes out of the cheapness of the polling process, and won’t be addressed in the NZ context. Proper polling is expensive, and NZ media won’t pay the premium required.

    • Gosman 2.4
      30 May 2011 at 12:35 pm

      Can anyone explain to me Mr Bradbury’s, seemingly warped, logic whereby someone is willing to commission fundamentally flawed polling that just reinforces positions that they have already taken.

      Surely if you are interested in finding out what the electoral impact of your policies are you would want your polling to be as accurate as possible not just regurgitating what you want to hear.

      • r0b 2.4.1
        30 May 2011 at 12:48 pm

        Use a little imagination Gossman. There are any number of reasons that a government might want to keep spreading false figures about how popular it is.

        However, in the current New Zealand context, I just don’t think they’d get away with it on a large, systematic scale.

        • Gosman 2.4.1.1
          30 May 2011 at 1:55 pm

          You lefties are an amazing bunch. Noone would want dodgy polling as it would mean decisions based on faulty data. Yet if you and that nut job lefty Mr Bradbury are to be believed that is exactly what the National party is doing.

          • Lanthanide 2.4.1.1.1
            30 May 2011 at 2:14 pm

            You don’t seem to get it, Gosman.
             
            It is possible to do two sorts of polling – private polling, which is what people really think and is never published. Then there’s public polling, which is where you come up with a perverted version of what people think (using whatever methods necessary) and then put it in all the mainstream media, thereby continuing whatever narrative it is that you want to promote.
             
            A polling company in the US got busted for this last year. Nate Silverman of fivethirtyeight.com (# of electoral college votes in the US) noticed that one polling company consistently skewed to the right in all it’s published polls. He did statistical analysis on it and showed that it had to be faked, and a few months later they somehow slipped up and their internal process was revealed, where indeed they were faking the data in exactly the way Nate predicted they were. If there was no benefit in producing skewed polling data, why was that company doing it?

            I’m not suggesting that level of corruption is going on here, by a long shot, but publishing skewed results clearly has benefits for those doing the skewing.

            • Gosman 2.4.1.1.1.1
              30 May 2011 at 2:15 pm

              And you have evidence that the mainstream polls in N.Z. are of this latest sort?

              Waiting………

              • Puddleglum …
                30 May 2011 at 10:06 pm

                Please read the last sentence in Lanthanide’s comment. He doesn’t have to provide you with any such evidence because he isn’t making that claim.

                • Gosman …
                  31 May 2011 at 10:46 am

                  Actually Lanthanide stated that he wasn’t suggesting that the LEVEL of corruption he mentioned was happening in NZ. The implication being that some sort of corrupt polling is taking place. Trying to weazle out of accusations like this isn’t very pretty.

                  • Puddleglum …
                    31 May 2011 at 1:26 pm

                    Gosman, you wrote:

                    “And you have evidence that the mainstream polls in N.Z. are of this latest sort?” (emphasis added)

                    As I read your phrase ‘of this latest sort’ I assumed you were referring to the ‘sort’ just detailed by Lanthanide. That is, the ‘sort’ carried out by an American polling company as mentioned in the paragraph beginning:
                    “A polling company in the US got busted for this last year.”

                    In that light, I took your request as being for evidence that such polling was being carried out here. Lanthanide explicitly pulled back from making that claim ‘by a long shot’.

                    Little did I know that, in your mind, ‘this latest sort‘ referred to NZ polling – and use of polling – that was at some unspecified, much lower LEVEL of corruption than the ‘latest sort’ exposed in the US, and to which Lanthanide had only made the most oblique reference to (and which he did not suggest was the ‘latest sort’ of polling here).

                    It just seems very odd to me that you would refer to current (and presumably historic) polling in New Zealand by the phrase ‘this latest sort’ in a discussion of ‘innovative’ polling techniques in the US.

                    No ‘weazling’ at all on my part – unless taking the most obvious meaning of your words generally amounts to unfair ‘weazling’.

                    • Lanthanide …
                      31 May 2011 at 2:48 pm

                      Gosman routinely displays deliberate mis-interpretation of most people’s comments on this blog, and is very sneering and condescending about it. So usually I just ignore him and let readers make up their own mind.

                    • lprent []
                      31 May 2011 at 3:24 pm

                      Echo of my approach (unless I feel like hitting something and have nothing better to do) , and based on the steadily decreasing numbers of replies he gets, probably most others as well.

          • r0b 2.4.1.1.2
            30 May 2011 at 2:18 pm

            Gosman – can you read at all? Two things.

            Yet if you and that nut job lefty Mr Bradbury are to be believed that is exactly what the National party is doing.

            (1) No – that’s exactly what I say is Not Happening.

            Noone would want dodgy polling as it would mean decisions based on faulty data.

            (2) National don’t give a damn about data. They are mostly driven by ideology and greed.

            • Gosman 2.4.1.1.2.1
              30 May 2011 at 2:25 pm

              Pure ugly biased opinion there r0b.

              Your dehumanising of anyone other than those of a left leaning persuassion is a big reason why the left is trailing so badly in the polls. Most people don’t like the arrogant holier than thou attitude of people on the left like you.

              • Colonial Viper …
                30 May 2011 at 2:26 pm

                Another Righty blaming others for what is actually their own mindset.

              • r0b …
                30 May 2011 at 2:32 pm

                Gosman you go about stating my views as completely the opposite of what they are. That tells us all we need to know about both your comprehension skills and your hate filled world view. You want “dehumanising” take a look in the mirror.

                • Gosman …
                  30 May 2011 at 2:34 pm

                  So pray tell r0b how you are not dehumanising people of a right leaning perspective when you state the following – “National don’t give a damn about data. They are mostly driven by ideology and greed”.

                  • r0b …
                    30 May 2011 at 2:42 pm

                    I don’t think you know what “dehumanising” means Gosman. Ideology and greed are very much human conditions. Unfortunately. And they very much drive the National Party. Ideology in the form of tax cuts, trashing savings schemes, and misguided spending cuts, when all the data shows that these are foolish things to do in a recession, and further privatisation, which has already done so much to wreck our economy. Greed in the feathering of their own nests, from tax cuts to their rich selves, and bum warming BMWs while they preach austerity to the rest of us. Ideology and greed all the way.

                    • Gosman …
                      30 May 2011 at 2:48 pm

                      This is your opinion r0b, albeit a strongly held one. However it still is attributing to those people on the left a set of core values which make them seem less than human.

                      I wouldn’t state that people on the left are mostly driven by their ignorance and envy even though I firmly believe those traits are prevalent in many of the comments people from the left make.

                    • r0b []
                      30 May 2011 at 3:00 pm

                      However it still is attributing to those people on the left a set of core values which make them seem less than human.

                      People on the right, Gosman, try and keep your rants tidy eh.

                      It’s a bit of a worry how quick you are to classify values as “less than human”. But if you’re not disputing the examples that I gave, I guess we’re done here.

                    • Gosman …
                      30 May 2011 at 2:57 pm

                      Sorry, that should be on the right not left in the first paragraph.

          • RedandBlack 2.4.1.1.3
            30 May 2011 at 11:35 pm

            Haven’t you watched Question Time in Parliament recently Gosman? The false polling is used as a weapon by Ministers supposedly answering questions. Not all of us were born yesterday!

    • Peter 2.5
      30 May 2011 at 5:20 pm

      Good call Bomber, keep telling it like is is!

  3. randal 3
    30 May 2011 at 11:58 am

    dont worry folks. stu dickinson will make sure ozzy wins the world cup and then keys and his mates will be dog tucker. singing on the road to gundagai as we ride off into the sunset.

  4. todd 4
    30 May 2011 at 1:02 pm

    Being that there’s a large difference between these reportedly “accurate” polls, makes me believe that there are inconsistencies in the way they’re undertaken. I would be far more likely to believe polling results if this difference was not so large. That margin clearly points to the fact that polling is being undertaken in a less than robust manner.

    There needs to be a set structure that ensures more accuracy and proper representation. This would in fact benefit those who believe people can be swayed through a hive mind mentality. It’s hard to say how many New Zealanders would vote just to back a perceived winner, but this is problematic in two ways. Firstly it misrepresents a misinformed public and secondly it does not allow those in power to perceive the public’s acceptance of their policies.

    There is a serious and unfortunate theme that is presently pervading the current political debate within New Zealand. The dissemination of information is being obfuscated. This allows a politician to lie to the public who does not have access to factual information. Much of the relevant information required to conduct a proper argument is being removed from websites. With the media being in the pocket of the Government, we can expect a very lop sided debate leading up to the next election.

    Only through proper public debate on policy with factual information can we hope to save our democracy.

    • Draco T Bastard 4.1
      30 May 2011 at 5:06 pm

      Only through proper public debate on policy with factual information can we hope to save our democracy.

      And there’s absolutely no way the RWNJs would allow that if they can prevent it. As they seem to be in power ATM I’m sure that they are doing the damnedest to prevent it.

  5. bomber 5
    30 May 2011 at 1:19 pm

    I agree that there is no systematic attempt to rig the polls, I am claiming that in a recession as steep as this, the right wings tools to gain political opinion have gone awry however the rise and rise of David as a supposed commentator on more and more platforms utterly unchallenged is an attempt to spin opinion rather than represent it.

    It is in his direct interests to sell these Polls as legitimate – the glaring falsehood of Len Brown supposedly neck and neck with John Banks by the Herald DigiPoll is simply ignored to promote polling methods that are not in fact representing legitimate opinion and are an attempt to simply manipulate.

    • r0b 5.1
      30 May 2011 at 1:31 pm

      the rise and rise of David as a supposed commentator on more and more platforms utterly unchallenged is an attempt to spin opinion rather than represent it.

      Well yes, quite agree. The media is more and more overrun by right wing propaganda, of which Nat hack David “billboards” Farrar is a prime example. But that’s a separate issue from the polling.

    • Gosman 5.2
      30 May 2011 at 2:01 pm

      Your jealousy of David Farrar’s success in getting his opinions on to the mainstream instead of poor old you knows no bounds doesn’t it Mr Bradbury?

      A friend of mine wants to know if you were the same ‘Bomber’ that was the roving reporter on Television back in the 1990’s with Petra Bagust on Ice TV. Nice to see you have been as successful in your media career as your contempories from that time.

      • onsos 5.2.1
        30 May 2011 at 8:19 pm

        Weak ad hominem. You fail.

        • Gosman 5.2.1.1
          31 May 2011 at 10:46 am

          Countered by an even weaker one of your own.

          • onsos 5.2.1.1.1
            31 May 2011 at 11:03 am

            Go to wikipedia, read what an ad hominem is, and then explain how I’ve conducted one. You conducted one when you attacked Bomber’s position on the basis that he was jealous.

    • queenstfarmer 5.3
      30 May 2011 at 2:59 pm

      Jealousy really is the ugliest trait

      Didn’t hear you complaining about Matt McCarten’s regular Herald column (in print and online). (and which IMHO is usually quite good, though I’d disagree 9 times out of 10)

      • Gosman 5.3.1
        30 May 2011 at 3:10 pm

        No apparently Matt McCarten’s is balanced by the ‘evil baby eating’ Deborah Doddington. The only way to balance David Farrar would be if some leftist blogger with a (smallish) media profile was to also get to post their opinion in the Herald. Hmmmm…. I wonder if Mr Bradbury has anyone in mind?

        • Colonial Viper 5.3.1.1
          30 May 2011 at 6:02 pm

          McCarten doesn’t try and represent his opinions as neutral and impartial without political leanings, you idiots.

          • Gosman 5.3.1.1.1
            31 May 2011 at 10:47 am

            So are you implying that David Farrar does attempt to portray himself as neutral.

            Do you have any evidence for this?

            • felix 5.3.1.1.1.1
              31 May 2011 at 11:01 am

              God, you’ve been back here two days and you’ve gotten to the bottom of your list already.

            • onsos 5.3.1.1.1.2
              31 May 2011 at 11:06 am

              DPF has a disclosure statement. Simply stating at the bottom of his entries that he is a National Party member and contractor would be quite straightforward. Instead I am expected to follow another link and read more of his waffle.

              He discloses crap about what he believes after his affiliations. I don’t care what he believes–he can have as many conflicts of interest with himself as he likes.

  6. Zaphod Beeblebrox 6
    30 May 2011 at 1:26 pm

    The TV1 and TV3 polls were reasonably close last time so why would you not disbelieve them?

    The bigger issue is the brain dead commentary which asserts that something is good because more people support it in polls. Apparently JK is a genius and PG a clutz because errr… one does better in public opinion polls.

    According to that logic UFOs must be real and shooting at refugees coming to Australia is a good thing because errr.. polls support it.

    Some journalists need to get an opinion and start using their brains.

    • Puddleglum 6.1
      30 May 2011 at 10:34 pm

      Agreed.

      This is the sort of focus on ‘competition by numbers’ that is to be expected in a society inculcated in ‘hollow’ market ideology. It is ‘hollow’ because of its pure worship of the process rather than what the process involves. It doesn’t matter what the ‘market’ throws up it is, by definition, ‘good’.

      Similarly, once politics is reduced to winning it is completely irrelevant what else happens as a result. Keeping on winning is the sole, surviving value. Every other value is not only just secondary, it is irrelevant.

      The psychologist Skinner famously argued that the only value, ultimately, was survival. Well, yes, but that fails to take account of the fact that we are never in that situation in the sense that there is no such thing as ‘ultimate survival’ – there is only ever survival via some other act. For humans, that translates as survival via some other value.

      You can’t survive simply by having a value to survive (whatever that contentless, hollow notion means). Typically, that contentless value actually manifests and gets fleshed out by rather humdrum, banal values such as short-sighted selfishness, deceit and the like.

      There are too many self-styled clever dicks seduced by this kind of absurd hyper-logic (Ayn Rand, Ludwig von Mises, Don Brash, Roger Douglas, etc.). They need to stop acting like 13 year old schoolboys hunched over their computers, believing they know better than the practical wisdom of millenia.

      They won’t stop, though – you see, like adolescent Machiavelli’s, they think they know the dark secret of the ages. Probably explains why they always become obsessed by ‘dungeons and dragons’-type games – and economics … and party politics. 🙂

    • McFlock 6.2
      30 May 2011 at 10:42 pm

      The TV1 and TV3 polls were reasonably close last time so why would you not disbelieve them?

       
      Because of the “floating voter”, who probably isn’t responding as much to surveys at this stage.
      I’m not sure how horizon did last election – were they even running? – but it’s quite probable that polls close to the election are more accurate than polls six months out, simply because people pay closer attention as decision time approaches.
       
       
       
       

  7. mickysavage 7
    30 May 2011 at 1:27 pm

    My favourite example is the 2005 Colmar Brunton poll taken just before the time of the election.  It predicted National winning 44 to 38.  Labour won 42 to 40, an 8% difference.  The figures are here.
     
    I suspect that polls are less reliable now because of the landline issue.
     
    Also my feeling on the ground is that a lot of people who did vote for them are peeved but this is not showing up.  Hopefully these polls will not sap the morale of the left although I suspect this is what the powers that be want them to achieve.
     

    • r0b 7.1
      30 May 2011 at 1:33 pm

      Yeah, which is why nervous nellies like McCarten and Trotter are doing the Left more harm than good. Bless their misguided wee hearts.

  8. bomber 8
    30 May 2011 at 1:38 pm

    @ Zaphod – I agree that the Polls have been able to be close in the past, but those polls were almost 3 years old now and we have suffered the steep recession which I am arguing has damaged the tools those pollsters used – and I point directly at the Herald DigiPoll the day before super city voting claiming Brown and Banks were neck and neck.

    • Zaphod Beeblebrox 8.1
      30 May 2011 at 1:45 pm

      Bomber- love your Fri night show on Stratos.

      Not sure the Auckland mayoral contest is a typical example. What happened there was a large turn out in the South and West of Auckland of people who nobody expected to take an interest, let alone vote.

      If you can guarantee that people will be similarly motivated and excited by Phil and Labour you might have a point, but I’m not sure that is that likely.

      • onsos 8.1.1
        30 May 2011 at 8:27 pm

        It should be easier to get accurate polling of a smaller population. A lot is known about the demographics of Auckland, and it should have been apparent that significant amounts of voters weren’t being represented.
        This wasn’t a deliberate blunder; it was a result of cheap polling process. This may be unavoidable. The polls need to be treated with disdain.

        • Zaphod Beeblebrox 8.1.1.1
          30 May 2011 at 10:29 pm

          The problem as I understood it was that turnout at this election was much higher than at any other recent Auckland LG elections.

          Everyone except for the Brown camp expected in-difference and dis-interest (in South Auckland in particular) which didn’t eventuate. Wondering if the pollsters will make the same mistake again.

          • onsos 8.1.1.1.1
            31 May 2011 at 9:15 am

            So why was this trend missed? This was the most striking fact from that election, and it was predictable, and it was missed. This is, incidentally, a systematic failing in NZ polls.

    • Gosman 8.2
      30 May 2011 at 2:01 pm

      What does the Herald Digipolls have to state about National politics?

    • mickysavage 8.3
      30 May 2011 at 5:14 pm

      One important factor that the polls did not recognise was that the Labour vote stayed home last time.  For instance in Manukau East, a Labour stronghold, turnout was 5,000 fewer votes than before.  I am sure they would have said “Labour” when asked but just stayed at home instead.
       
      I do sense the current polls are not correct.  As long as Labour can battle the sense of malaise they cause and get people out to vote the election could be interesting.

  9. ak 9
    30 May 2011 at 3:03 pm

    r0b: I agree that they become self-reinforcing, and are used to shape opinion almost as much as they reflect it.

    Exactamundo, well understood by the parties of Marketing, as evidenced by the massive repetition by both the privately-owned media and right-wing commenters – wish I had a cent for every “blah blah blah…. look at our big polls!”. The gravy issue is just how much the polls shape opinion, and that’s something that can’t be measured.

    And don’t forget the “refusal rate” (“politics? nah piss off I’m having tea”) reported here once as >70%. Who would we imagine most represented in that group?

    Another wee point: it’s not “the polls” that have been “roughly good predictors” of actual results – rather you’re describing the final week’s polls: who’s to say the previous years’ polls were all either bogus and/or inaccurate and themselves influenced the actual final vote. The self-fulfilling prophecy scenario. Remembering always that it’s slolely the relatively politically apathetic “swingers” who we’re talking about in all these discussions: anyone who’s sat through multiple meeting situations knows full well how participants are regularly and relatively easily “swayed” to the “winning” position, even from previously-held strong positions.

    Even then, the 2005 example given by mickey above (and bomber’s Lenslide job) disprove the “good predictor” theory as far as recent events go – not to mention the consistent righty bias of the Brolmar. The Horizon results certainly seem to gell better with “the vibe”: wish it weren’t so, but it’s hard to believe Brash got no significant lift from his recent poison.

    Despite all that, you’re dead right r0b; lot’s more work needed on the ground from Labour and us all. Mining, Mt Albert and Lenslide showed the way, let’s look forward once again to “blah, blah, blah…look at our big vote!”

  10. Bored 10
    30 May 2011 at 4:39 pm

    There is only one Poll of any relevance: that from the booths of South and West Auckland (compounded by the greater Porirua, Aranui area votes) on election night. If the left can get these people to the polls on their behalf no 15% gap in good old middle class polling land will make an iota of difference. The election needs to be fought down south and out west on the issue of NACT attempting to impoverish Pasifikans /”white trash” further.

  11. RedandBlack 11
    30 May 2011 at 6:06 pm

    What we can’t ignore is that mainstream media in NZ have been taken over by the Right with two exceptions – Sunday Star Times and Campbell Live. Even RNZ has succumbed – how else would Farrar and Matthew Hooton get any air-time?
    Chris Trotter has sold out – a far cry from the activist he once was.
    Media is so dumbed down that TVNZ can state that National has the numbers to govern alone on a skewed Colmar Brunton poll.
    What is happening in NZ today is similar to what happens in Russia. Russia has virtual political parties. New Zealand has virtual polls and virtual commentators.

    • Campbell Larsen 11.1
      30 May 2011 at 6:06 pm

      er, Sunday Star Times and Campbell Live exceptions!!!???

      I don’t think so.

      • RedandBlack 11.1.1
        31 May 2011 at 10:52 pm

        Perhaps I was being over-generous. Talking about Right-wing media, this week’s Listener magazine has Key and English on the cover explaining the budget to the ‘scrupulously fair’ and ‘balanced’ Jane Clifton – cheap PR trick. I call it.
        In The Press Chris Trotter is not much better, quoting some stupid poll about the supposed popularity of DonKey and basing the whole article on that. Is that his genuine opinion or is he writing to order? Bring back John Minto!

  12. Huginn 12
    30 May 2011 at 6:53 pm

    The 28th May letter from Horizon outlining the methodological differences made the point that polls do influence outcomes when the Electoral Commission uses them as one of its criteria in allocating money to fund election broadcasting.

    Horizon made the point that Digipol’s methodology over represents support for larger parties and under represents support for minor parties.

    So to answer Gosman, a political strategist may well consider commissioning fundamentally flawed polling as a pre-emptive attack on minor parties like NZ First and the Greens.

  13. todd 13
    31 May 2011 at 10:57 am

    More Inaccurate Polling

    http://thejackalman.blogspot.com/2011/05/more-inaccurate-polling.html

    There’s been some debate recently concerning political polling within New Zealand. Many are questioning discrepancies within the research and believe that we’re not seeing an accurate representation of public opinion. This was no more apparent than in last nights TV3 News item with the right wing Duncan Garner reporting on National’s Kiwisaver cuts.

  14. Alexandra 14
    31 May 2011 at 1:53 pm

    I’m not sure whether this issue is part of the above debate as I dont have time right now to read all the comments.

    Ive been phoned a number of times recently by colmar brunton polsters and its apparent that great care is taken in targeting a particular kind of demographic. First you are screened to determine whether or not you a suitable to respond to the substantive questions. As soon as I identify my age, gender and ethnicity, I’m eliminated! So how can those polls be accurate in determining general public sentiment, if that is what they serve to do?
    Should we lie about who we are and manipulate the polls in order for the results to be more truthful?

    • Lanthanide 14.1
      31 May 2011 at 2:53 pm

      It’s because the polling companies have decided that they need to get x number of people of a certain age, gender and ethnicity, and y people of a certain other age, gender and ethnicity. This is so that their sample better reflects NZ demographics, and allows them to draw statistical inferences from the small sample size, which is much cheaper than getting a very very big sample size.

      If they just interviewed whoever answered the phone, they might end up (through pure bad luck) with a sample that is 750 men and 250 women, even though NZ society is much closer to 500/500.

      So it’s only natural that when they phone you up, if you fall outside of the remaining pool for participants (eg you’re a male and they already have 500 men) then they won’t interview you.

      • Alexandra 14.1.1
        31 May 2011 at 6:29 pm

        Oh OK, so they have already filled their quota of middled aged maori women by the time they get to me, everytime!

      • Anne 14.1.2
        31 May 2011 at 8:54 pm

        My personal experience with phone polls has been almost solely around age groups. That is, they wish to speak to someone between 35 and 45 years or someone under 25years. My own age group (at the upper end of the middle age bracket) seems to be decidedly out of favour!

        Unless it’s because there are lots more middle-aged people with landlines. I’ve long suspected there are a disproportionate number of older people being polled who tend to be more conservative in their politics.

        • Armchair Critic 14.1.2.1
          31 May 2011 at 9:22 pm

          You can always lie about your age, Anne – I’ve been known to if I want to partake in a survey and would not otherwise be eligible.
          You can also tell them whatever you want about how you have voted, or plan to vote. For example, there are pollsters out there that think I will vote National this year, even though I swore never again after 1990.

          • wtl 14.1.2.1.1
            31 May 2011 at 9:41 pm

            I don’t get it. Why would you want to overstate the level of public support for National if you don’t like them?

            • Armchair Critic 14.1.2.1.1.1
              31 May 2011 at 9:55 pm

              1. I can then say, with confidence, that the polls are not accurate.
              2. Pride comes before the fall, hopefully for National.

          • Anne 14.1.2.1.2
            1 June 2011 at 7:19 pm

            Yep.. done that Armchair. Voted NZ1 once . Can’t remember what I gave as reason :wink:. Trouble is: as you grow older you seem to become more honest. Is it because you forget to lie?

  15. Penny Bright 15
    31 May 2011 at 3:55 pm

    In my considered opinion – the only polls really worth considering are polls of voters – ELECTION results.

    National should be more than a little concerned with the recent Howick by-election results?

    70% of voters didn’t bother, and the Howick Ward of the Auckland Council covers the Botany and Pakuranga electorates – both ‘safe’ National seats?

    Yes – this was a local not central government election – but we are talking essentially the same voters.
    (It’s hardly top secret the connections between C & R and National? )

    Think you will find that there are a lot of National Party / C & R voters who still get power bills every month, and know from their own experience that the Rogernomic$ reforms that introduced ‘competition’ into the ‘natural monopoly’ of electricity supply just duplicated resources and caused massive power increases for ‘customers’.

    (Interesting that in less than 3 months – went from 128 votes in the Botany by-election to 1491 in the Howick by-election!

    Seems some traction is being gained on the following issues:

    “NO RATE$ INCREASE!”

    “CUT OUT THE CONTRACTORS!”

    “$UPERCITY = SUPER RIP OFF!”

    FYI – I intend to stand in Epsom for the General Election, as an ‘Independent’, continuing to ‘make a fuss’ on the following pro-transparency; anti-corruption, anti-privatisation ‘IT’S TIME TO ROLLBACK ROGERNOMIC$’ issues….

    Look forward to promoting the ‘partial rate$ revolt’ against the 3.7% Auckland Council rates increase in the Epsom electorate.

    The Auckland Council /CCO books are NOT open – ‘full and accurate records’ of independent contractors are NOT being maintained. In my considered opinion, if a huge ‘scalpel’ were to be applied to the consultants and private contractors ‘blubber’ carrying out core council functions – potentially hundreds of million$ could be sliced off rates bills.

    (I have lodged a formal complaint with the Chief Archivist for this alleged breach (s 17(1) of the Public Records Act 2005.)

    Penny Bright
    http://waterpressure.wordpress.com

    • todd 15.1
      31 May 2011 at 5:00 pm

      Excellent Penny, you have our support. Give Act a right bollocking! You might even get some coverage from the media.

    • Zaphod Beeblebrox 15.2
      31 May 2011 at 6:47 pm

      I know he his not a friend of yours Penny but the Cetacean on Citizen A last week claims he is aware of 20K contracts being handed out by Auckland Council untendered. Be interested to see if he has the evidence.

  16. Tiger Mountain 16
    31 May 2011 at 4:13 pm

    Good onya Penny, your ‘brand’ is on the rise! Politics is about real life for some of us. We get out there and do stuff as well as blog. Dare I mention it but 30th anniv for ’81 tour coming up. That kind of spirit will be needed again very soon in NZ.