Conservative

Labour and the polls

When Andrew Little was first elected to the Labour leadership last year, everyone I knew seemed a little dumbfounded. No one, whatever their political stripes, thought him a good choice, with their reasons generally ranging from his apparently humourless personality to his union credentials. Labour had doomed itself, was the general consensus.

Of course, Mr Little then failed to make a right hash of things. In fact, with his “Cut the crap” soundbite, he got a fair bit of positive press coverage, and seemed to have united many Labour Party doubters behind him.

Nonetheless, despite that good initial run of form from him, the mood from the streets of Gisborne, Rotorua and wherever else I wandered remained circumspect. People expected Little to come a cropper sooner or later (perhaps sooner rather than later), and they were hardly likely to switch allegiance to (or back to) Labour until the party had shown it could offer some degree of basic competency.

On Sunday, we had the first television poll of the year: TV3’s Reid Research poll. There was good news for Labour – it was on 29.1%; up 3.5% (or up 4% on its election night result) and close to the near-respectability figure of 30%. And 55% of voters thought that Mr Little was “potentially a better match for Mr Key than his predecessors”.

Leaving aside the usefulness or otherwise of including the word “potentially” in that last polling question, the results really aren’t great news for Labour. For a start, National was up 5.3% to 49.8% (or up 2.8% on their election night result). Labour’s rise in support didn’t come from National…

It’s the post-election summer, and until this frenetic last week, politics has been off everyone’s agenda. With no media exposure, the smaller parties have suffered:

  • The Greens were down 5.1% to 9.3% (or down 1.4% from election night), having had little more than bad publicity since the election after their lacklustre 10.7% showing.
  • NZ First is on 6.9%, down 1.9% on its election night result (although down just 0.2% on the pre-election TV3 poll).
  • The Conservatives are on 2.7%, down 2.2% from the last TV3 poll and down 1.3% from the election.
  • Internet Mana are on just 0.6%, down 1.4% from the last TV3 poll and down 0.8% from the election.

Labour and National have basically just profited from the usual post-election lack of exposure that the minor parties tend to suffer. Labour’s 29.1% leaves the party no closer to governing. National’s back in governing alone territory.

And Andrew Little’s preferred Prime Minister rating was just 9.8%. It’s not a dreadful debut, but it’s still 2.5% less than the terminally disliked David Cunliffe was polling.

Really all that can be said about the TV3 poll results are that there is still more than two and a half years to run until the next election, so it’s early days yet. Plenty of time for Little and Labour to build on an error-free Parliamentary term. Or plenty of time for an implosion.

Advertisements

Rebirth of the Poll of Polls

So, how did my Poll of Polls do? Pretty rubbish really… The rapid rise of NZ First and the Conservatives during the last half of the campaign didn’t come through in my results, and there certainly seems to be something systemic about the Greens’ ability to fall short of their poll results come election day.

So, I’ve been playing around with the numbers, and have messed with my algorithm to produce what should (hopefully) be a more accurate beast. The changes involve further front-loading of the weighting of new polls (so that the Poll of Polls responds more quickly to meteoric rises a la NZ First and the Conservatives), updating the in-house polling bias offsets, and introducing industry bias offsets (to hopefully deal with issues such as the systematic overly high poll results for the Greens or the lower on average results for NZ First, compared to election day results).

The Poll of Polls is therefore reborn, all ready for today’s Roy Morgan poll. If Labour needed any further evidence that the public think the party is in a hopeless state of disarray, this is it. The party hits a new Roy Morgan low of just 22.5%. National also slumps, hitting 43.5%.

The Greens are the big winners, hitting a record high of 17.5% – cold comfort, given their relatively lacklustre election result. Of the remaining minor parties, NZ First is on 7%, the Maori Party is on 2%, ACT is on 0.5%, United Future is on 0.5%, the Conservatives are on a record high of 5%, and Internet Mana is on 1%.

So how does the new (hopefully) improved Poll of Polls look?

National: 46.1% (-0.9% from its election result)

Labour: 24.6% (-0.5%)

Greens: 11.8% (+1.1%)

NZ First: 8.3% (-0.4%)

Maori: 1.2% (-0.1%)

United Future: 0.3% (+0.1%)

ACT: 0.7% (nc)

Internet Mana: 1.6% (+0.2%)

Conservative: 4.2% (+0.2%)

Based on those percentages, the parties are predicted to win the following number of seats:

National: 59 (-1 from its election result)

Labour: 32 (nc)

Greens: 15 (+1)

NZ First: 11 (nc)

Maori: 2 (nc)

United Future: 1 (nc)

ACT: 1 (nc)

Internet Mana: 0 (nc)

Conservative: 0 (nc)

Given National’s drop in the Roy Morgan, and the Greens’ outlier of a result, it’s not surprising to see National lose a seat to the Greens. Whether the Greens can hold anywhere near their Roy Morgan support in other upcoming polls remains to be seen…

The Right bloc sits on a total of 61 seats, compared to 47 for a Labour, Greens and Internet Mana alliance, meaning National could continue to govern with the support of both United Future and ACT.

Where to from here for National?

If John Key wants to have a stab at a fourth term as Prime Minister, there’ll be no one in the party to stop him. He’s weathered the Dirty Politics and Moment of Truth storms, and come out the other side with an increased majority.

Now it’s time for a clean up. Jason Ede has already resigned, which is perfect timing for National. An announcement prior to the election would have looked like an admission of guilt, just prior to people walking into the polling booth. This way, it’s lost in the honeymoon maze, and when the House returns to sit, the opposition will have lost another line of attack.

You’d hope that National’s leadership has learned its lesson from the Dirty Politics saga, and will keep people like Cameron Slater at bay. National may have romped home, but Brand Key has lost a touch more of its shine in the process. National’s result wasn’t necessarily as much an endorsement of John Key’s charms as a rejection of the state of the Left.

And hopefully, National MPs (and prospective MPs) lower down the food chain learn from the reaction within National to the Dirty Politics claims regarding Slater and Lusk’s involvement in the Rodney electorate selection process. If anyone finds out you’ve contracted Slater or Lusk to run interference for you, you’ll hopefully be toast.

Of course, the big issue for National, as they seek re-election in 2017, is the same one that kept them awake at night over the last three years – coalition partners. The election results for ACT and United Future were risible. National will give them roles in this new Government though, partly as a reward for six years of loyalty, partly in the vain hope that they might against all odds surge again in popularity and offer National more assistance at getting over the line in three years time.

Likewise, the Maori Party will be offered a role again too. Te Ururoa Flavell has been very clear that the Maori Party cold work with both National or Labour. National will be keen to keep Flavell onside.

But what if that’s not enough? What if ACT and United Future remain unappetising minnows, and Team Key needs a few more seats next time? Does National build up the Conservatives in the hope that they’ll supplant NZ First?

Once the honeymoon fades, Steven Joyce and the rest of the strategy team will undoubtedly be pondering what needs to be done to ensure a victory in three years time.

The deconstruction – what went down

So, in the end it wasn’t even close. Unless the special votes are dramatically out of kilter with the votes counted on election night, National has the numbers to govern alone.

The worse-case scenario now for National is that they lose a seat to the Greens, meaning that National would need one of either ACT or United Future to pass legislation. It’s not such a terrible worst-case for the Nats – both ACT and United Future are entirely dependent on National for their continued survival; they wouldn’t be giving National too much stick. Besides, as Graeme Edgeler writes at Public Address, if the special votes are distributed in the same proportions as in 2011, there’ll be no change to the makeup of Parliament.

So how has National managed to defy the laws of electoral gravity, while Labour plumbs new depths, and the minor parties are all left licking their wounds? For me, there are five main issues.

Firstly, the public were happy with our current economic stability. On The Nation, just before National released its tax “policy”, Bill English couldn’t highlight a single new idea that National would bring to the table to spur economic growth. It turns out the public weren’t too unhappy about that. People presumably looked at the economic mess that exists in the US, the EU and across the ditch in Australia, and thought that things were actually pretty good here. Labour’s ideas might have sounded interesting, but would they work? Do I trust Labour to mess around with Kiwisaver rates, and what will the effect be on my weekly take-home pay?

Secondly, for almost three years (and another three before that, if you include the Goff years), Labour has presented itself as a chaotic pack of self-absorbed in-fighters, too busy playing identity politics and sticking the knife into opposing factions to give a damn about Middle New Zealand. Labour may have stayed on message with grim determination during the actual campaign, but by then it’s a bit late. Staying on message for six weeks cannot outweigh more than two and a half years of self-mutiliation. The public had already made up its collective mind that Labour were a pack of muppets.

Thirdly, the public were happy with our current political stability. Put simply, Internet Mana scared the hell out of people. I’ve lost track of the number of people I’ve talked to who weren’t necessarily National Party fans, but who wanted Kim Dotcom’s hands nowhere near the levers of power. Those people likely voted National. The Left may point to ACT or Colin Craig, and ask what’s the difference? Well, part of the reason Colin Craig didn’t get the endorsement he wanted was that National’s polling indicated National would take a 2-3% hit if they gave Craig a seat, so there’s not necessarily much of a difference there. And ACT simply isn’t seen as a threat these days; it’s been dependent on National for so long that it’s been politically neutered. Internet Mana, on the other hand, was a frightening unknown; a Frankenstein mix of hard-left activism and big money.

Fourthly, Dirty Politics largely wiped out any emphasis on policy. Yes, the policy was out there, and Dirty Politics almost didn’t feature during the leaders’ debates, but a huge chunk of the election campaign was lost to it. The minutiae of the allegations were largely lost on the public. Judith Collins made an easy villain, and she resigned – case largely closed. Dirty Politics wasn’t seen as a reason to change a Government.

Finally, the Moment of Truth. It was the moment where Kim Dotcom took his credibility out behind the bike sheds and hit it with both barrels, making himself (and the Left, by association) look like idiots. But it also inspired a large dose of parochialism in the dying week of the campaign. “I’m not going to be lectured to by a bunch of foreigners” – it was a phrase I heard rather a lot of, in various permutations. On The Nation yesterday, David Farrar told Lisa Owen that his polling for National showed a jump in National’s support following the Moment of Truth.

Kim Dotcom and his Moment of Truth may have handed John Key those final few seats he needed to attain the ability to govern alone.

 

So, predictions…

I’ve been keeping track of the polls with my Poll of Polls (final update here), but of course polls technically aren’t prediction devices. They ask the question, “If an election were held today/tomorrow”, and are therefore only so useful when it comes to predicting what people will do in a few days time.

Likewise, Poll of Polls’ are generally fairly slow at adjusting to sudden events. They help cancel out statistical noise, but sometimes when a party shoots up in the polls it’s not statistical noise; the party actually is significantly more popular than it was the previous day, week or month.

The rise of NZ First and the Conservatives is a case in point. The final pre-election polls from each of the five main polling companies shows a spread of 6.6% to 8.4% for NZ First (an average of 7.6%), while my Poll of Polls has them on just 6.3%, outside the spread altogether.

Likewise, the final polls for the Conservatives show a spread of between 3.3% and 4.9%, while my Poll of Polls has them on 3.3%, at the very bottom of the spread.

Then there’s the perennial issue of whether the polls are inherently biased. Are they missing important swathes of the voting population, resulting in fundamentally skewed results? My Poll of Polls adjusts each poll based on how far above or below the industry average that polling company is. It doesn’t adjust for whether the polls are inherently out in relation to election results., largely those results can change quite markedly from election to election.

There are a few bias-adjusted predictions out there. Over at the Dim-Post, Danyl McLauchlan Poll of Polls applies a significant downward adjustment to National, and a significant upward adjustment to NZ First (there are other adjustments, but those are the big ones). I think his adjustments are too large, but there you go… I guess we’ll soon know just right or otherwise he is…

Danyl has given his predictions for five parties, heavily couched with 2% bands:

  • National 42 – 44%
  • Labour 22 – 24%
  • Greens 13 – 15%
  • NZ First 7 – 9%
  • Conservatives 5 – 7% (although he further couches his prediction by noting that the recent controversy over the resignation of Colin Craig’s press secretary might drop the Conservatives below 5%).

And Gavin White has published his bias-adjusted predictions for the parties he has “good data” for:

  • National – 45%
  • Labour – 26%
  • Greens – 11%
  • NZ First – 9%
  • Maori – 1.3%
  • ACT – 0.8%
  • United Future – 0.6%

My gut feeling prediction?

  • National 46%
  • Labour 26%
  • Greens 12%
  • NZ First 8%
  • Conservatives 4%
  • Maori – 1%
  • United Future – 0.2%
  • ACT – 0.5%
  • Internet Mana – 1.9%

Now let’s see how wrong I am come Saturday night!

Poll of Polls update – 19 September 2014

It’s time for the final pre-election Poll of Polls update! We’ve had the last Herald Digipoll and Fairfax Ipsos poll results this morning, so we’re good to go. (If Roy Morgan suddenly publish a three-day poll this afternoon, then bugger ’em!)

In the Herald Digipoll, there’s a minor fall for National, down 0.4% to 48.2%. Labour rise 1.3% to 25.9%, while the Greens fall back 0.5% to 11%.

NZ First are sitting pretty on 8.4% (up 0.3%), while the Conservatives fall of the pace, down 0.5% to 3.3%.

For the remaining minor parties, the Maori Party is on 1.1% (up 0.4%), Internet Mana is on 1.0% (down a large 1.3%), ACT is on 0.5% (up 0.2%), and United Future is on 0.2% (up 0.2%).

In the Fairfax Ipsos poll, National slump a massive 5.1%, but that still leaves them on a relatively respectable 47.7%. Labour meanwhile rises handsomely, up by 3.7%, but that leaves them on a still disappointing 26.1%. The Greens lose ground, down 1.0% to 12.0%.

NZ First is on 6.6% (up 2.2%), while the Conservatives miss out on the 5% threshold in yet another poll, hitting 4.5%, despite increasing 0.9%.

The Maori Party is on 0.9% (down 0.1%), United Future fails to register, ACT is on 0.3% (down 0.4%) and Internet Mana is on 0.9% (down 0.5%).

So here’s how the Poll of Polls ends up:

National: 47.6% (-0.1%)

Labour: 25.9% (-0.1%)

Greens: 12.6% (-0.1%)

NZ First: 6.3% (+0.2%)

Maori: 1.0% (+0.1%)

United Future: 0.1% (nc)

ACT: 0.5% (nc)

Internet Mana: 1.9% (-0.1%)

Conservative: 3.3% (+0.1%)

Based on those percentages, the parties are predicted to win the following number of seats:

National: 60 (nc)

Labour: 32 (nc)

Greens: 16 (nc)

NZ First: 8 (nc)

Maori: 2 (nc)

United Future: 1 (nc)

ACT: 1 (nc)

Internet Mana: 2 (nc)

It’s uneventful final update, with no movement in the seat count, and just minor movements in the party vote stakes. Labour drop to a new low, but that’s almost par for the course these days.

That means that the Right bloc remains on a total of 62 seats, compared to 50 for a Labour, Greens and Internet Mana alliance, meaning National could continue to govern with the support of both United Future and ACT.

Poll of Polls update – 18 September 2014

We’ve had the final pre-election One News Colmar Brunton poll tonight, so it’s the second to last Poll of Polls update before the election, with the Herald Digipoll and Fairfax Ipsos polls due out tomorrow (although you can already find the Digipoll results if you look on Wikipedia – that’s what happens when you release the percentage changes and expect that no one will do the maths…).

National drop 1% to 45%, leaving them well and truly in danger zone territory. Labour remains static on 25%, while the Greens fall 2% to 12%.

NZ First comfortably waltz back into Parliament with 8% (up 1%), as opposed to the Conservatives, who may have gone up by 0.4%, but still remain below the 5% threshold on 4.4%.

For the remaining minor parties, the Maori Party doubles its vote from 0.8% to 1.6%, Internet Mana is on 1.8% (up 0.4%), ACT is on 0.6% (down 0.6%), and United Future fails to register (down 0.2%).

So here’s how the Poll of Polls looks now:

National: 47.7% (-0.2%)

Labour: 26.0% (nc)

Greens: 12.7% (nc)

NZ First: 6.1% (+0.2%)

Maori: 0.9% (nc)

United Future: 0.1% (-0.1)

ACT: 0.5% (nc)

Internet Mana: 2.0% (nc)

Conservative: 3.2% (nc)

Based on those percentages, the parties are predicted to win the following number of seats:

National: 60 (nc)

Labour: 32 (-1)

Greens: 16 (nc)

NZ First: 8 (+1)

Maori: 2 (nc)

United Future: 1 (nc)

ACT: 1 (nc)

Internet Mana: 2 (nc)

In the party vote stakes, National continues to fall, dropping another 0.2%. They’ve lost half a per cent in just 4 days. Meanwhile, NZ First continues to rise, breaking past 6% for the first time this year.

NZ First’s party vote gains are enough to see it take a seat off Labour, who therefore lose the seat they won off Internet Mana last update.

That means that the Right bloc remains on a total of 62 seats, compared to 50 for a Labour, Greens and Internet Mana alliance, meaning National could continue to govern with the support of both United Future and ACT.

Labour candidates’ destiny out of their hands

Consider the Super 15 (or whatever name the competition is currently going by), as the final round of the regular season arrives. Most teams don’t have a chance at qualifying top of their conference, but there’s still a chance of getting through in one of the remaining spots. But various results have to go their way. Team X must lose to Team Y by 23 points. Team M must draw with Team Q. Their destiny is no longer in their own hands.

Come Election Night, there’s a few sitting Labour MPs who might well be in a similar position. This site’s Poll of Polls currently has Labour on 26.0%, with 33 MPs. Let’s assume that Labour gets 33 MPs on Saturday, and look at who might be in or out.

First, some assumptions. Carmel Sepuloni will win Kelston, and Jenny Salesa will win Manukau East. One’s in a new seat, and the other’s a new candidate, but they should romp home.

There are some relatively marginal seats, but it’s likely than not that Damien O’Connor will win West Coast Tasman, Iain Lees-Galloway will hold Palmsterson North, Trevor Mallard will win Hutt South, Stuart Nash will win back Napier, and Tony Milne will win back Christchurch Central.

If those are the only marginal results that go Labour’s way, then Raymond Huo would be the cut-off point on Labour’s effective list. Carol Beaumont will be gone, as would Ruth Dyson (who isn’t on the list, and is dependent on winning Port Hills in the face of unfavourable boundary changes).

But what happens if a few more close races go in Labour’s favour, with Adrian Rurawhe winning Te Tai Hauauru and Peeni Henare winning Tamaki Makaurau? Well, Kelvin Davis and Raymond Huo won’t be returning. And if Ruth Dyson wins Port Hills? Then it’s sayonara to Moana Mackey.

Attempted new entrants Priyanca Radhakrishnan and Tamati Coffey must have initially thought their respective list positions of 23 and 30 were pretty good. With Labour’s current polling though, Ms Radhakrishnan is certainly no shoe in, and even if Trevor Mallard was to lose Hutt South, Adrian Rurawhe and Peeni Henare were to lose their Maori seat campaigns, and Stuart Nash was to fail in Napier, Tamati Coffey would still only be the next cab off the rank.

List MPs such as Sue Moroney, Andrew Little, Maryan Street and Moana Mackey will be hoping that the Conservatives get 4.9%, therefore bumping up the effective Labour Party vote share.

Quite a few on-the-cusp Labour MPs may be spending their Saturday night hoping that their colleagues fail in their electorate challenges…

Poll of Polls update – 17 September 2014

Two new polls out today – this afternoon’s Roy Morgan, and this evening’s 3News Reid Research poll.

In the Roy Morgan poll, National rise to 46.5% (up 1.5%), while Labour slumps 2% to 24%, and the Greens drop an even larger 2.5% to 13.5%.

It’s a great poll result for NZ First, who rise to 8.0% (up 2%). It’s their second best result from any major polling company this year, beaten only by the 8.1% they achieved in the latest Herald Digipoll.

For the remaining minor parties, the Maori Party hits 1.5% (up 1%), United Future rises from zero to 0.5%, ACT loses half of their support, dropping from 1% to 0.5%, and there’s no change for Internet Mana and the Conservatives, on 1.0% and 3.5% respectively.

The 3News Reid Research poll, meanwhile, is a very different beast. National achieve their worst result this year from any major pollster, hitting 44.5% (down 2.2%).

It’s not great news for Labour though, who also drop – down 0.5% to 25.6%. It’s yet another mid-twenties result – in fact, it’s the twentieth major poll in a row in which Labour has failed to get to 30%.

The Greens are up 1.4% to 14.4%, while NZ First has another great result, hitting 7.1% (up 1.2%).

The big talking point though is the Conservatives. They’re on 4.9% (up 0.2%); so close to the 5% threshold it’s frightening. Of course, Reid Research is the only polling company that has them above 4%, so it remains to be seen whether they’re right and everyone else is wrong…

For the remaining minor parties, Internet Mana are on 2.0% (up 0.3%), the Maori Party is on 1.1% (down 0.2%), ACT is on 0.1% (down 0.2%), and United Future is on 0.1% (up 0.1%).

So here’s how the Poll of Polls looks now:

National: 47.9% (-0.3%)

Labour: 26.0% (-0.1%)

Greens: 12.7% (+0.1%)

NZ First: 5.9% (+0.2%)

Maori: 0.9% (nc)

United Future: 0.2% (+0.1)

ACT: 0.5% (nc)

Internet Mana: 2.0% (nc)

Conservative: 3.2% (+0.2%)

Based on those percentages, the parties are predicted to win the following number of seats:

National: 60 (nc)

Labour: 33 (+1)

Greens: 16 (nc)

NZ First: 7 (nc)

Maori: 2 (nc)

United Future: 1 (nc)

ACT: 1 (nc)

Internet Mana: 2 (-1)

In the party vote stakes, the biggest mover is National, falling 0.3%. That puts them at their lowest result since early June. Labour also fall (again) though, and they reach (another) new Poll of Polls low.

With National and Labour predictably falling during the election campaign, it’s the rise of the minor parties, with another Poll of Polls high for the Greens, NZ First and the Conservatives.

Seat-wise, last update Labour loses a seat to Internet Mana, by the most minuscule of margins. This update, again by the most minuscule of margins, they win it back.

That means that the Right bloc remains on a total of 62 seats, compared to 51 for a Labour, Greens and Internet Mana alliance, meaning National could continue to govern with the support of both United Future and ACT.

Out in space, no one can hear you scream

From the sidelines of relevance, ACT has been screaming and waving its hand in a desperate bid to be noticed. The bold (some may use less charitable descriptive language) policies have been coming thick and fast recently – abolishing the Overseas Investment Office, getting rid of the Resource Management Act, arming shopkeepers…

They’re designed to grab headlines, to give party leader Jamie Whyte a few life-giving gulps of publicity oxygen. If publicity is supposed to get you votes, it doesn’t appear to have working for ACT. Following their campaign launch and OIO and RMA announcements, the last Colmar Brunton had them on a year-long high of 1.2%. Unfortunately, they could then only manage a combined 1% from the latest Digipoll and Ipsos polls.

Things are only going to get worse this week. Tonight, Kim Dotcom unveils his Moment of Truth. Whatever Dotcom and Glenn Greenwald have to say tonight, be it bunker busting explosion or damp squib, the media won’t be talking about much else. Certainly, the upcoming War of the Documents between Greenwald and Key won’t be a simple day-long skirmish.

Minor party policy will well and truly be taking a back seat. ACT and United Future might as well just concentrate on Epsom and Ohariu respectively, and give up on the rest of the country. Given that Colin Craig has no hope of winning an electorate seat, Dotcom’s Moment of Truth might very well be the final nail in the coffin of the Conservatives’ run for the 5% threshold.

In fact, the coming week is likely to be a policy-free vacuum full stop. Which means that for the opposition, they’d better hope that whatever Dotcom and Greenwald have to say is compelling and easy to understand.

David Cunliffe, Russel Norman and Meteria Turei had better get used to the idea that for the next four days, the only soundbites they’ll be giving will be on spying. For the Greens, that focus might still pick them up a few votes, given their long-term activism against our spy agencies. For Labour though, Dotcom’s Moment of Truth may prove catastrophic. All eyes will be on John Key and whatever he may end up declassifying. With four days of Key v Dotcom & Greenwald, David Cunliffe and Labour Party policy won’t get a look in.

The stakes are high. If Dotcom’s “proof” that Key knew about Dotcom before the raid isn’t watertight, Internet Mana will be a laughing stock. Likewise, if the Greenwald v Key debate about mass surveillance gets lost in a maze of paperwork and semantics, Labour and the Greens can probably give up on being anything other than bit players for the final week of the campaign. The only winner is likely to be Winston Peters, sweeping up the disillusioned.

The fortunes of Labour, the Greens, and Internet Mana are now firmly anchored to two men. Roll on 7pm…