Monday, October 29, 2012

Newspoll Provides More Evidence That The Abbott Elastic Has Snapped

NOTE ADDED LATER: This is an old article - if you are looking for the current Newspoll because this one comes up when you Google for Newspoll, you won't find it here! Sorry!  For my analysis of the post-spill Newspoll of late March 2013 see here.

 Advance summary:

1. The current Newspoll strongly confirms the pattern observed in a previous piece that the relationship between Tony Abbott's approval ratings and two-party polling is shifting. 

2. There is also some evidence that Labor's polling is better than it should be given its own leader's current ratings.

3.This is all consistent with either the view that Abbott's leadership is not impacting on Labor as much as it once did, or else the view that the impact of Abbott on Labor's standing was exaggerated by many in the first place.

------------------------------
 I won't normally make a habit of flooding this site with posts whenever a federal pollster releases a poll, but I will comment on individual polls when I think the results are significant, especially in the context of an ongoing theme, the interplay between 2PP and leader ratings.

In last week's article about Opposition Leader approval ratings and party standing I noted that an inverse connection between Tony Abbott's popularity and the Coalition's standing in the polls, which has existed throughout the Abbott leadership, has now started showing signs of wear and tear.  The Opposition's lead has been shrinking without Tony Abbott's ratings improving.  The suspicion advanced in my article (not for the first time either) is that most of Labor's troubles have been self-inflicted and that Abbott is not only fairly irrelevant to Labor's 2PP but perhaps even dragging the Coalition's down.

Here is a revised version of my graph of the relationship between net satisfaction for Tony Abbott and the Coalition's 2PP since the last election.




The large pink dots are the last three Newspolls.  One of them (the one before this) is surrounded by blue dots and within the normal Abbott 2PP/netsat range).  However the other two (at (-30,0) and (-28,0) are way outside the normal Abbott range.

Newspoll has been notorious for "bouncing" lately, although the long term history of Newspoll shows an average of nearly 2 points 2PP difference between successive polls, which is bound to increase when there is an actual change in voter sentiment unfolding.  So how does the pattern look if we smooth the data?  Here's the same graph again but both axes are smoothed.  My smoothing method for Newspoll is a running average of the last four polls with the current poll weighted double.


On this graph, we're currently at the lowest of the three pink dots and we are further outside the blue zone than we have been before; Abbott has his second-worst rolling average netsat ever and the Coalition rolling 2PP is now 51.8, and will go down again next poll unless they get at least 53.

One way of looking at this is that by normal Abbott standards, this unpopularity level should mean a Coalition 2PP of 55.3.  Another way of looking at it is that with this rolling 2PP, Abbott's rolling netsat should be -10 rather than -27.

Recent polls by other pollsters that have canvassed approval ratings have been quite consistent with this emerging picture.

But it is not only Abbott for whom the graphs are showing something odd.  Here is the rolled-data graph for Gillard:

 (Ignore the blue data point at the far right; that was her first and was a "honeymoon bounce" rating that couldn't be rolled with prior data since there were none.)

The current rolling average is the highest of the pink dots, and the pink dots suggest that Labor is currently polling a bit better than Gillard's popularity (which, let's not forget, drives over 80% of the 2PP variation) predicts. 

Those who believed the depths of Labor 2PP to which Labor sunk mid-term were down to Abbott's leadership need to face the evidence that either he is nowhere near as effective now as he once was, or else it was mostly never his doing in the first place. 

Furthermore, it's widely accepted that Labor's collapse going into the 2010 election was down in part to the dangers of poll-driven government (ie governing, and even removing the leader, based on short-term impacts on party standing.)  What is not so widely accepted is that during this term Labor, whether through strategic intention, haplessness, a Green gun to the head or sheer lack of a viable alternative, has not usually pursued such a path on policy, tactics or leadership.  It is the Coalition that has displayed poll-driven politics on all these scores throughout this term, continually taking an attitude that if they are up 56-44 then all is right with the world.

And, while it is certainly the case that the Coalition is still polling a few points better on rolling average than a party with such an unpopular leader usually would, that's not much consolation given the historical competition and the very favourable hand the Coalition holds in terms of things to attack an incumbent government over.  Indeed, a few months before he was rolled, Alexander Downer had similarly bad netsats but a better 2PP than Abbott does now. 

If the Coalition wins the next election by even a small margin - and Labor's ability to lose federal elections should never be underestimated - it is going to all look like genius.  But as Peter Brent has often noted, it is the usual habit of commentators to resort to what I call "analysis by result" * - whatever the winner does is brilliant, whatever the loser did was terrible, even if most of these actions actually had no impact at all.  If the Coalition do win the next election under Abbott, it may well be in spite of his leadership and not because of it. 

(* Adapted from the equivalent chess term "annotation by result")

No comments:

Post a Comment

The comment system is unreliable. If you cannot submit comments you can email me a comment (via email link in profile) - email must be entitled: Comment for publication, followed by the name of the article you wish to comment on. Comments are accepted in full or not at all. Comments will be published under the name the email is sent from unless an alias is clearly requested and stated. If you submit a comment which is not accepted within a few days you can also email me and I will check if it has been received.