Australian elections – are polls bad, or does media misuse them?

Scott Morrison and his National Coalition winning the Australian election is being reported as a shock, in part due to polls predicting a loss.  Are polls a waste of time? Or does media put too much weight on imprecise indications of how people might vote?

I keep saying that at best polls are an approximate indication of how people may vote in the past, and can in no way predict accurate election results in the future. Polls have well known statistical margins of error, but media reporting on them seem to largely ignore this.

Perhaps more accurate ways could be found to predict election results, but I think that a media obsession with trying to predict what will happen in the future is aas much a problem as polling methods.

RNZ – Australia election: Why polls got it so wrong

It was predicted to be the federal election Labor simply couldn’t lose, but after last night’s surprise Coalition win, the opinion poll may struggle to stand the test of time.

Experts say cost cutting and technological change in the polling process is leading to many inaccurate and misleading suggestions.

Nearly all polls predicted Labor leader Bill Shorten would have an easy win with a 51:49 lead over Prime Minister Scott Morrison on a two-party preferred basis.

I dispute that. Polls generally ask something like ‘if an election was held today who would you vote for?’ – perhaps some polls ask ‘who will you vote for on election day?’ but i have never seen that.

And a 2% gap is well within margins of error, which are usually around 3-4%.

51% with a 3% margin of error means there’s a 95% chance of the result being between 48% and 54%.

49% with a 3% margin of error means there’s a 95% chance of the result being between 46% and 52%.

So there is quite an overlap.

In fact, for three years the polls had picked the Opposition to take government.

Again I dispute that. Over the last three years polls tried to measure who people might vote for in the week or two prior to the poll being published.

They are usually whole country polls. Elections in non-MMP countries like Australia and the USA can be decided in just a few key swing electorates or swing states. \being swing electorates they have a history of impressionable swing voters.

Election campaigns are carefully planned to try to change crucial votes right up until election day. Polls are not designed to examine how people mat change their mind at the last minute.

I obviously have ideas about who to vote for in the weeks and days before an election, but I don’t decide for sure until I vote. There must be others who do similar. Polls can’t get inside our heads.

So why exactly were the polls, as ABC political editor Andrew Probyn put it last night, such a “shambles”?

Former Newspoll boss Martin O’Shannessy blamed the flawed forecasting on the fact that many people’s telephone habits have changed.

“Not everybody has a landline and the numbers that are published are incomplete.”

That might be a part of the problem – but that doesn’t address the ‘trying to predict the future’  misrepresentation of polls.

Polls can only be approximate.

I think that media trying to use polls as precise predictors of future voting is the biggest problem here.

Media activists versus Bridges

Everyone can see that Simon Bridges has been struggling to impress as National’s leader. Many have said so, and not just those who wish that any National leader and their party will struggle.

It’s the media’s job to report what they see.

But it shouldn’t be the media’s job (or rather some who are presented as political journalists) to try to get any leader dumped, or to try to promote an alternative leader.

However this is what seems to be happening. And to an extent the media have the power to make it happen.

John Armstrong:  Media script requires Bridges to end up as dog tucker

The media have proclaimed Simon Bridges to be dog tucker. Having issued that decree, the media will do its darnedest to make sure he does become exactly that – dog tucker.

That is the ugly truth now confronting Bridges in his continuing struggle to keep his leadership of the National Party intact and alive.

It is unfair. Some pundits had made up their minds that Bridges was the wrong person to lead National within weeks of him securing the job. Those verdicts were quickly followed by bold predictions that it would not be long before he was rolled by his fellow MPs.

No account was taken of the difficulty of taking over a political party which has been thrown into the irrelevance of Opposition after having called the shots from the Government benches in Parliament for nigh on a decade.

It is not media bias at work here, however. When the media hunts as a pack – as is the case with Bridges – it is colour blind.

It is not fussy about where it feeds. It is not fussy whether the victim comes with a blue or a red tag. If you doubt that just ask Andrew Little.

Or David Shearer or David Cunliffe.

Armstrong is right – the media don’t care what colour the blood is, they smell it liken sharks and go in for the kill.

That the media are so rabid is simply the consequence of the adversarial nature of politics. The media are consumed with what is going wrong rather than what any government or Opposition party might be getting right.

I think this is only partly right. The long established adversarial nature of politics is part of the reason – but that combines with two more recent trends – the desire (and need) for clocks online, and also the rise in the level of personal involvement, advocacy and activism by journalists. Some of them are far from detached observers and reporters. Some want to be moves and shakers.

The hunt is constant for inconsistency, gaffes, blunders, infighting and so on. Negativity rules, OK.

Makes ‘better’ headlines than positives.

Despite its efforts, the media claim few scalps by their devices alone. They are instead vultures hovering over the road-kill offered up in the preferred prime minister ratings in what is now a sporadic number of polls.

That there are now only two news organisations commissioning such voter surveys – and at three-monthly intervals – means discerning a trend can be virtually impossible.

To draw conclusions from the surge in backing for Labour and decline in support for National registered by the Newshub Reid-Research poll is folly.

It was hardly a surge for Labour – it was a notable but one off change from their last poll (ignoring any other polls) nearly a year ago. And it was barely a decline for National in the current political context of being in opposition and taking into account margins of error.

Likewise the preferred prime minister ratings. That Collins has overhauled Bridges was used to reinforce the notion that Collins is now a viable candidate for the leadership — and that Bridges is not.

In part media are making the Collins conflict – they have been boosting her as an alternative by giving her exposure with every poor poll for Bridges. They have also effectively chosen her over all other National MPs.

Once you are deemed to be a loser by the media, that becomes a mindset which is near impossible to erase.
The loser falls victim to a feeding frenzy – and there can only be one outcome from that.

So the media end up getting their scalp and headlines – for a short while, before looking for the next victim.

Labour leaks targeting Bridges

There have been a series of leaks of internal information obviously designed to damage Simon Bridges and National.

This began with the odd expenses leak just a few days before the information was due for public release, followed by the onslaught from Jami-Lee Ross as the now ex-National MP self destructed. There have been further anonymous leaks of historical information that look suspiciously like a continuation of that attack.

There has also been what looks like a Labour campaign to discredit Bridges and destabilise National heading into the holiday period.

Leaked UMR polling information has progressed from whispers to journalists to drip feeing of poll graphics. I posted on this one yesterday –UMR polling history – which notably was monthly polling with the last result from October, so without the latest poll. One could presume someone is only able to get old data, or the November poll didn’t fit the hit.

There is also a word cloud floating around – Stuff reported on it here How public view Simon Bridges – that was purportedly ‘sent to corporate clients in late November’ and has just popped up. This also indicates it is October data – from the time of the Jami-lee Ross saga, so an out of date targeted hit on Bridges.

Ex Labour staffer Neale Jones, now working for a ‘public affairs company, specialising in Government Relations, Strategic Communications and Campaigns’, keeps tweeting a stream of criticisms of Bridges and National. Whether that is personal or part of Strategic Communications and Campaigns is not clear.

And The Standard has a steady diet of anti-Bridges/National posts. Over the past week:

Mostly this is preaching to the converted, and several authors are involved, but it looks like they have more interested in damaging the Opposition than promoting the Government.

Over the same period there are three posts on Labour/Government bills.

Will all of this have any overall effect? It’s hard to say, but even though there has been a string of media ‘opinions’ from political journalists dumping on Bridges the consensus is that a leadership challenge would be unlikely with National polling higher than Labour (apart from the leaks of cherry picked UMR polls.

In the meantime Jacinda Ardern and Labour keep polling reasonably well – but news of Government progress has not been prominent. Perhaps that’s why there is more focus on attacking National.

Political polls for 2018

Political polls for the year haven’t shown any drastic changes, with Labour and National swapping the lead a few times after Labour had risen to be competitive late last year after the election.

I presume there will be no more political polls for 2018. Colmar Brunton (for 1 News) are the only ones left doing polls, and they have just published what will be their last one for the year.

Reid Research (Newshub) did just two polls this year, in January and May. Roy Morgan have up given doing New Zealand polls. Their last poll was in November 2017.

Labour looked dire mid 2017 but Jacinda Ardern’s leadership turned things around for them enough for them to  be able to form a government, thanks to NZ First.

NZ First have remained in the MMP danger zone, peaking on the 5% threshold but dropping as low as 2.4% (in May).

After polling mostly in the 10-15% range in the first half of last year Greens dropped drastically after the Turei fallout, and through this year holding their support just over the threshold in the 5-7% range. So their support has halved from the support they got for most of last term.

It seems normal for coalition support parties to struggle to maintain support.

After the latest poll Ardern was criticised for claiming that Labour “finishing the year stronger than we started it”, but she is correct, sort of, by a small margin and she is comparing two different polling companies.

Reid Research did an unusually early poll in the political holiday period 18-28 January, and had Labour on 42.3%. In May they had Labour on 42.6%.

Colmar Brunton’s last poll (24-28 November) had Labour on 43% (rounded so could have been as low as 42.51% or as high as 43.49%). However Colmar’s first poll of the year (10-14 February) had Labour at 48% so Labour have dropped back from that Colmar high.

Ardern also said “polls do move around a bit these are all still within the margin of error” –

We can only see trends from Colmar – here are Labour’s results for the year.

  • 10-14 February 48%
  • 7-11 April 43%
  • 19-23 May 43%
  • 28 Jul – 1 Aug  42%
  • 15-19 October 45%
  • 24-28 November 43%

The 48% for Labour looks to be a polling outlier – it could have been accurate at the time, but Labour settled in and remained in the low forties for the rest of the year. While they will be disappointed to be trailing National this is a fairly solid result for them, considering their pre-Ardern polling had them dropping in the twenties. Colmar had them trending down to 24% in July 2017.

National’s results from Colmar this year:

  • 10-14 February 43%
  • 7-11 April 44%
  • 19-23 May 45%
  • 28 Jul – 1 Aug  45%
  • 15-19 October 43%
  • 24-28 November 46%

They were behind Labour in February and in October (affected by the Jami-Lee Ross mess) but this is remarkably consistent for a party in Opposition, and with new leader Simon Bridges (since 27 February) who is struggling to make a mark.

Looking at the Labour and National polling for the year there is little in it, and little significant change in most polls.

Media have tried to make big stories out of their polls, but the reality is quite mundane.

I think we have a real problem with how polls are reported. Obviously media try to get bang for their bucks – polling can be expensive – but they usually make mountains out of mole polls, often blatantly misrepresenting what individual polls mean.

Media try to make each of their polls look like some sort of mini election, which is nonsense. They can only be approximate indicators of support, and the year after an election most of the people care little about politics most of the time.

If media were doing proper journalism they would report on the political polling without sensation and misrepresentation. And mostly that would be (and should be) quite boring.

How should the media get value for the money spent on polls? Perhaps they should also poll on things of real public interest at the same time, and make their big stories about that.

1 News blew that opportunity in the last poll. They did ask a one-off question – Should Simon Bridges boot Jami-Lee Ross from Parliament using waka jumping law?

The results of that mean nothing (and were inconclusive, with 31% saying they didn’t know). Most people have moved on from one MP self-destructing – actually most people probably took little notice when the media were going hard out with headlines.

1 News would probably like to encourage National to chuck Ross out of the waka (that would be out of parliament, they have already chucked him out of the party) because that could be headlined as a sensational political somersault or something.

Rather than aiming for short term headlines 1 News could do a really public service (they are a public media company after all) doing a series of meaningful polls on issues that really matter to people, but it would take months if not years to get a return on their investment. They seem too obsessed with short term ratings and clicks.

So I expect more of the same form polling next year, another non-election year. It’s a shame we are so poorly served by media who do polling, but I don’t see that changing.

Something worse has become prevalent – online polls run by media. They are cheap, and nasty, very unreliable so they are of no useful purpose.

Northcote by-election polls

Early voting has opened in the Northcote by-election. ‘Election day’ is 9 June.

There have been claims from both Labour and National about polls leading into the voting period, however these are internal polls commissioned by the parties with no details made public, so should be viewed with scepticism. I’m posting about them here to put it on the record, so claims can be compared to results.

The Daily Blog on Thursday:

The latest internal polling by Labour astoundingly suggests that Labour could beat National in Northcote!

I’ve just seen the latest polling for the Northcote by-election and it’s much closer than we thought.

Going into this race we knew we were underdogs – and, if I’m honest, it looked like we had a huge mountain to climb to even get close given National’s huge majority.

But the latest polling we received yesterday reveals there are just a few points in it. There is a chance Labour could do well on June 9.

Labour have just sent this out to their Northcote support team, and if true could be one of the biggest upsets in a  safe National electorate since Winston took Northland!

Labour has emailed not just members but apparently it’s whole email list (someone I know in Dunedin with no connection to Labour got one).

Referring to a tweet that it had been posted on The Daily Blog:

We will just have to wait and see what the result is. A lot may depend on party ‘get out the vote’ organisation and voter motivation, which can be low for by-elections like this with little but party egos at stake.

Typically single electorate polls can have quite small sample sizes so have higher than normal margins for error.

And by-elections with nothing much riding on them can have very low turnouts. Turnout comparisons in Mount Albert (which Jacinda Ardern won early last year):

  • General election 2014 – 36,922 votes
  • By-election 25 February 2017 – 13,649 vote
  • General election 22 September 2017 – 38,760 vote

These are the by-election candidates in Northcote:

Candidate Name Party
BERRY, Stephen ACT New Zealand
BIDOIS, Dan National Party
CHEEL, Tricia Democrats for Social Credit
HALBERT, Shanan Labour Party
JAUNG, Rebekah Green Party
KOLONI, Kym Independent
LYE, Jeff Aotearoa Legalise Cannabis Party
WALSH, Liam NAP

Trump improved in polls, a bit

Donald Trump’s approval ratings have been better for most of the last two months (February, March) than they have been over the last twelve months, but they are still negative:

It’s difficult to know why there has been an improvement this year – it will be a combination of factors. Getting tough on trade may be having an effect, and recent recovery may have been influenced by Trump joining the international condemnation of Russia. There is no obvious sign of the Stormy Daniels affair affecting support.

RCP Poll Average trends:

Interestingly the latest poll result is from Rasmussen which has Trump on -8% overall approval (45-53). In late February they had his approval slightly ahead of disapproval (50-48) but his approval has slipped gradually since then.

Despite the improvement Trump is still well below the approval of Obama and GW Bush at the same time into their eight year terms, as shown by FiveThirtyEight:
It’s difficult to predict how things will go from here.

The tax cuts will have pleased many. Time will tell how they work out – Trump has just agreed to a huge spending bill and the US is significantly increasing deficits and debt levels, and that will impact over time.

Much may depend on Trump’s trade ‘wars’ and the economy – the US sharemarket had improved significantly right through last year but ironically dropped steeply at about the same time Trump’s approval ratings improved at the start of February.

Jobs were a big deal in the 2016 election campaign. Success or failure there can only be judged over time.

Trump seems to have had some success in his confrontation with North Korea, and has agreed to meet with Kim Yong Un, but that has not been organised yet – the north Korean leader has just visited China and has said he will denuclearize.

The Middle East could go any way – if Trump can move the region towards peace he will win a lot of credit, but Syria and Afghanistan are still looking very dicey.

Trump’s White House administration has always struggled to get staff and has turned over a lot of high level staff – this could reflect the rush to fill positions when unexpectedly winning the presidency, with a gradual sorting out of who fits his leadership style and policy preferences – or it could be rats jumping ship. Trump’s manner or firing and humiliating people who fall out of favour may discourage potential replacements.

And the Russian election interference issue is still simmering away without a clear idea whether that will damage Trump or his family, but he must have concerns given his attacks on the inquiry and the FBI. This could all fizzle out at the Trump level, or it could blow up big time. The jury is still out on that, with a lot of the investigation details still under wraps.

Evidence against TOP

The Opportunities Party have promoted their policies as evidence based. From About on their website: TOP takes a long term, evidence based view.

However now we are down to the business end of the campaign evidence seems to have flown out the TOP window.

A few days ago on Newshub: Gareth Morgan blames landlines for poor polling, claims he’ll win 5-10 percent

“When I ask the question in the town hall shows I do every night, ‘ Hands up those who’ve got a landline, it’s 10 or 15 percent,” the Opportunities Party (TOP) leader told The AM Show on Thursday.

“What’s wrong with these polling companies? I think we’ll be somewhere between 5 and 10 percent. I’ve said it from day one.”

Where’s the evidence? TOP has a big budget, if they wanted evidence they would have done their own polling. I think it’s quite likely they have done their own polling, if so it is not evidence they want publicised.

Cut Your Hair: The evidence says TOP have no hope

TOP pride themselves on being an evidence-based party. So it behooves us to examine the evidence behind Gareth Morgan’s suggestion that TOP have a real chance of winning representation in Saturday’s election.

Question: Has any party ever achieved what TOP is trying to achieve?

Short answer: No.

Long answer: Only one party has ever won representation under MMP in New Zealand without a sitting electorate MP from a sitting party. That sole exception is ACT, who had several prominent former Labour and National cabinet ministers. That happened in the first MMP election, when everyone and their mum voted minor party.

Not many parties have won representation under MMP in New Zealand, whether through the 5% threshold or local seats. Only one MP has ever won representation for a party that didn’t have an MP elected in 1996 for one party or another: Hone Harawira, for Mana.

Most of the small parties that have won representation have done so via a local seat (Māori, Mana, Progressive, United, ACT, and NZ First have all coat-tailed in). Only 7 parties have ever reached 5%: National, Labour, the Greens, NZ First, ACT, the Alliance, and United Future. The last three have all failed more times than they’ve succeeded and have basically shriveled away to nothing (or, worse, to David Seymour). Scores of parties have failed to reach 5% OR a local seat: the Conservatives, Christian Heritage/Coalition, Legalise Cannabis, Destiny, Outdoor Recreation, Future, etc.

The latest public polls (that use a variety of polling methods):

  • Listener Bauer Media Insights 1-5 Sept: 2.2%
  • 1 News Colmar Brunton 2-6 Sept: 1.9%
  • Newsroom-SSI 4-6 Sept: 2%
  • Roy Morgan 28 Aug-10 Sept: 2%
  • Newshub Reid Research 6-11 Sept: 1.6%
  • 1 News Colmar Brunton 9-13 Sept: 1.6%
  • Horizon Research 9-14 Sept: 2.3%

Evidently TOP look like getting nowhere near the 5% threshold.

So they have done their own polling. It shows them very likely to come up short.

Morgan will know that if they don’t look like getting close to 5% many voters will prefer to vote elsewhere rather than risk ‘wasting their vote’.  Hence the bullshit about the polls being wrong.

Question: Might the polling be wrong?

Short answer: Anything is possible, but TOP reaching 5% would require polling error on an unprecedented scale.

Morgan and Sean Plunket ranting and abusing on Twitter won’t change things.

It’s not just history and the polls that are against TOP. Others have tried Donald Trump’s tactic of being bellicose and abusive and complaining about the polls – in particular Winston Peters, and NZ First has slumped over the last two months in the polls.

Question: Is this a good year for a minor party to achieve the never-before-achieved?

Short Answer: No—on current polling this will be the worst MMP election ever for minor parties.

It looks like it will take a major game changer for TOP to get close to or beat the threshold, and they are running out of time.

Question: Could TOP win a local seat?

Short answer: There is no evidence to suggest they will come close to winning any local seat. Morgan might have had a chance, but he isn’t standing in a local seat.

That Morgan is targeting the polls and the threshold (without any evidence) supports this. TOP dabbled with targeting the Ohariu electorate a couple of weeks ago but that effort seems to have fizzled.

In some ways TOP have been impressive. Their evidence based approach to developing solid policies has been very good. Morgan has impressed sizeable crowds at campaign meetings.

But TOP has been shut out of small party debates. And they have failed to attract enough positive media attention. Morgan and Plunket have also been too cranky on Twitter and possibly elsewhere in social media.

Yesterday Plunket tweeted a challenge:

A bizarre approach.

It must be frustrating to have put so much time and money into their campaign, but making up shit about polls looks desperate and not based on any evidence.

Moaning about polls is almost certainly not going to change the game and suddenly boost support for TOP. Morgan might be better trying a different last gasp approach.

It’s sad to see another new party beaten by the ridiculously high threshold. Parliament could benefit from a different approach and some fresh ideas and MPs. But facts are facts, and TOP look like failing.

Poll trends

Polls are useful indicators, albeit backward looking. Trends are also worth keeping an eye on, but they can disguise sudden shifts.

David Farrar tweeted on a post at Kiwiblog:

This is quite misleading. I don’t know whether this is deliberate or not but the timing of this is questionable. August polls are already out of date. Farrar is National’s pollster.

There has been three public polls in September that add a lot to knowledge of poll trends.

Opinion polling for the New Zealand general election, 2017 is more up to date, and much more informative.

This shows that the dramatic Labour upswing has been sustained in September, and National support is diving.

It also highlights the Green dive in support, and shows that so far there is no sign of the  much touted late campaign improvement for NZ First.

 

The NZ First poll jump fallacy

Does NZ First always get substantially more votes in an election than they poll? No, there have been a number of variations over the last few elections.

This sort of claim at The Standard is common:

NZF always get 5% additional votes at the Election cf to the polls so they are probably tracking at 16-17% which would appear to be about right at this stage ?

I’ve often seen journalists claim that NZ First get more votes than they poll. Sometimes selective poll results are used to try to justify the claims.

But the fact is that NZ First polling results and trends compared to election results has varied markedly over the last few elections. And there have been varying factors involved.

In 2008 the trend remained quite flat for NZ First – election result 4.07%:

https://en.wikipedia.org/wiki/Opinion_polling_for_the_New_Zealand_general_election,_2008

In the months leading up to the 2008 election NZ First were embroiled in controversy over donations to the party. While Serious Fraud Office and Police investigated. Peters stood down from his ministerial roles (he was found not guilty of illegal wrongdoing).

In 2011 NZ First were all over the place, election result 6.59%

https://en.wikipedia.org/wiki/Opinion_polling_for_the_New_Zealand_general_election,_2011

Always the political opportunist Peters benefited from the controversy over the ‘tea tapes’ involving John Banks and John Key late in the campaign.

In 2014 there was a bit of a late upswing – election result 8.66%:

https://en.wikipedia.org/wiki/Opinion_polling_for_the_New_Zealand_general_election,_2014

This wasn’t a huge rise. In March 2014 NZ First had poll results up to 7% and through the year they often got 6.0-6.5%.

In August to December 2012 NZ First had poll results from 1.8% to 7%, often getting 5-6.5% (8 times in that range).

Every election differs, especially for smaller parties who can hardly be seen during the term and then get some attention during the campaign.

Some pundits have suggested that this election has similarities to 2005, except with major party roles reversed. National under Don Brash’s leadership had recovered from an abysmal 2002 low and went very close to beating Labour under Helen Clark.

A lot of late campaign focus then was on the two largest parties, and while NZ First ended up calling the coalition shots and ending up in government (remember the ‘baubles of power’) they lost ground late in the campaign.

In 2005 there was a late drop – election result 5.7%:

https://en.wikipedia.org/wiki/Opinion_polling_for_the_New_Zealand_general_election,_2005

In June and August NZ First had been getting up to 10-11%.

Peters may find an issue that resonates over the next five weeks and NZ First may get more votes than they have been polling. But they could just as easily stay flat, and a controversy could see them slide like the Greens just have.

Many different things could happen.

Labour are currently resurgent and could suck support away from NZ First.

Voters could fear a Labour+NZ First+Green coalition and rally behind National.

TOP may have a late surge of ‘stuff the others’ vote at the expense of NZ First.

Age and the rigours of having to campaign in an electorate and nationally may catch up on Peters – there are signs of strain showing. He could have a health scare.

Shane Jones may do or say something stupid (he’s been quiet lately) and scare voters off NZ First.

Nicky Hager could launch another book.

Cameron Slater may score a vital hit against National (he’s been trying hard enough, largely in vain).

Jacinda Ardern could be found wanting in election debates and voters may desert Labour, some to NZ First.

Election campaigns inevitably throw up surprises, and that can mean opportunities for smaller parties like NZ First – and can also deprive them of oxygen.

One thing is certain – a significant number of voters make late decisions about how they vote, and this means there can be significant shifts in support – as there have been over the last month.

Making presumptions based on selected past poll trends is futile.

Challenges for National

Despite the crisis faced by Labour, National still have challenges of their own heading into the election campaign.

Few expect them to do as well as last election when they just got a slim majority with 47.04%, trimmed to a whisker when they lost the Northland by-election.

Three recent polls:

  • UMR (Labour’s internal poll): 42%
  • Colmar Brunton: 47%
  • Newshub/Reid Research: 45.1%

I think the UMR poll is a bit older and tends to favour Labour rather than National, but these polls suggest that National have a challenge avoiding having to negotiate with Winston Peters based on all of these results.

BUT anything could happen now as a result of Labour’s torpidity followed by turmoil.

This could throw everything wide open if voters desert Labour and switch to the Greens or NZ First. Or voters may prefer the safety of an uninspiring but well known National plus a few small support parties again.

National still have the advantage of having a good relationship with several small parties still, while the Labour-Green Memorandum of Understanding was effectively thrown out the window by Metiria Turei and now Andrew Little seems to have smashed any window of opportunity for Labour.

So National has been helped by the turmoil on the left, but they have to do more to help themselves. Arrogance and complacency are probably their biggest challenges to overcome, but they could do with a much stronger approach from Bill English as well.

If they cruise then NZ First and the Greens make pick up the bulk of the votes thrown up in the air by Labour.