Poll of Polls: Plus Ça Change

Many polls. Much excitement. Little change.

We have updated the Pundit Poll of Polls, and it shows that the last few months of political manoeuvring, scandals, by-elections, and so on has not had any marked impact on the overall state of play. National remains in the driver's seat and on cruise control.

We estimate that National holds around 52% support, a figure it has hovered closely around since March. Labour has around 33% support, little changed since the last election, and very static since August. The consistent difference between the two larger parties hovers around nineteen points, right in the middle of its 17-22 point range during 2010. And the difference between the right-leaning and left-leaning parties as blocs, currently around 13%, has only varied within a narrow range (10-13 points) since May.

It isn’t like there has been nothing going on in politics during the year. There have been plenty of scandals and other items of prurient interest for political junkies. The 2010 Budget contained one of the biggest overhauls to our tax system in living memory. And the politicians’ responses to the tragedies in the South Island drew sustained headlines as well.

The parties seem to be stalemated because no consistent pattern has emerged in these events. For every Pansy Wong there was a Chris Carter. For every Key speech at Pike River there was a Brownlee debacle in parliament. And for all the people excited at the prospect of the Budget’s tax reform package, there appear to be just as many feeling ambivalent about its ultimate impact on their wallets. Without a consistent pattern in the events, there has been no catalyst for a media narrative about one team in the ascendancy and another on the decline. Without a narrative such as that, it is very hard to change the minds of citizens focused on things other than politics.

Among the smaller parties, the Greens have recorded a decline in recent weeks, which (while only 1%) accounts for about a seventh of their support base. Nonetheless they remain very comfortably above the 5% MMP threshold, despite what a solitary Colmar Brunton poll has recently suggested.

ACT’s slow steady slide from around 1.8% in July to around 1-1.2% in December represents over a third of their earlier supporters. Should this trend solidify, the value to National of giving Rodney Hide a good run in Epsom may only be one right-leaning MP – one National MP or two ACT MPs. Is that marginal gain, in an election where National will feel relatively comfortable about remaining in government, worth the price of having to put up with Rodney for three more years?

Also notable is one story that wasn’t – it turns out rumours of New Zealand First’s rise from the dead have been greatly exaggerated. Two reputable polls put New Zealand First at 4.5% in the second half of 2010, leading to a rash of coverage predicting Winston Peters’ Phoenix-like ascent from the ashes. The other thirteen polls – which attracted much less coverage of New Zealand First – showed this conclusion to be entirely premature. New Zealand First moved very little in our Poll of Polls in the second half of 2010. In July, their estimated support was 2.3-2.7%, in December it ranged from 2.5-2.7%. This is why we do polls of polls – reading polls one at a time can lead people to make all kinds of dramatic, exciting, and utterly wrong conclusions.

(We did also make one change to our Poll of Polls methodology. We no longer assume Jim Anderton will win Wigram, due to his impending retirement. This change means that the Progressives’ 0.4% support now translates into no seats rather than one overhang seat.)

All this stagnation has made Labour’s job even harder in the upcoming election year. For the left to have a legitimate shot at the Treasury benches, they need to cut the right-leaning bloc’s lead from thirteen points currently to three points on election day. Any less change than that makes a left-leaning coalition nigh-impossible to construct, and even with this level of change a right-leaning coalition would be much more likely to form. To bring about a ten-point swing like this would require Labour and the Greens to win over about 100,000 current National supporters while losing none of their own the other way. They have around 300 days to do it.

There is international precedent for large election year swings like this, for example in the recent Australian election. But such swings are very rare. Given the Poll of Polls’ history over the past year, there is little to indicate lighting on the horizon here.