Monday, November 9, 2015

EMRS: Liberals Rebound With Abbott Dead Weight Gone

EMRS: Liberal 48 Labor 25 Green 20 Ind 7
Interpretation: Liberal 49 Labor 29 Green 17 Other 5
Result of poll if election held now: Liberal Majority Government (probably 14-7-4)
Aggregate of all state polling: Lib 13 ALP 8 Green 4 (next most likely is 13-9-3)

A new EMRS poll of Tasmanian state voting intentions has been released and this one shows a massive eight-point shift to the Hodgman Liberal government since the previous poll in August. (Also see the trend tracker.) Really not a great deal has happened at state level since the August poll and I would interpret as much of that shift as isn't just random sample noise as being mainly down to federal factors.  The removal of Tony Abbott as Prime Minister has led to a general improvement in what little state Coalition polling has been seen around the country since the switch.  In the case of Tasmania, this is the Hodgman Government's best result from this pollster since a very similar result just after its election, and suggests it would retain its majority with one or maybe two seat losses.

EMRS polling tends to skew to the Greens and "Independent" and against Labor, so I interpret the poll as a little less dire to Labor than it seems.  All the same, the state swing is on paper only enough for Labor to pick up one seat in Braddon and one in Franklin.  In the case of the one in Franklin, a 1.3% swing is required, and while this poll points to about a 2% statewide swing, the Liberals would go into the election with three Franklin incumbents to Labor's one (compared with two each last time).  Thus I don't think Labor would regain that seat based on this poll.



I thought the result for the Liberals in the last EMRS was lower than their actual standing, and so I doubt they have really gained eight points on the back of the Turnbull shift (the federal government itself has so far gained six and a half.)   Also the current Turnbull boost is hardly any predictor of the next state election, which would probably be held in the middle of the federal government's second term if the federal government is re-elected.  All the same, it is better for the state government to be polling such numbers now than not to be polling them, and it is consistent with the idea that a cause of the state government's modest EMRS polling through its term so far was drag from a disliked federal regime.

That is not to suggest Abbott was really that much on the nose in Tasmania (most state-level federal polling was showing a muted swing compared to the rest of the country) but it could be that EMRS was picking up more of an effect than other pollsters.

Preferred Premier

The sees Will Hodgman with a very large lead (56-19) as preferred Premier, his greatest so far.  This should be seen in the context of the federal Turnbull bounce and the general advantage to incumbents in preferred-leader polling, but it's likely there is more to it than just that.  It would be interesting to see approval-rating polling for the leaders as the results might be unflattering for the Opposition Leader.  At the least, this result could reasonably be seen as a sign that after over a year and a half in the job, Bryan Green hasn't become a popular alternative leader and is generally seen as a stop-gap in the role.

EMRS have in fact been polling personal approval ratings - I was polled by them a month and a half ago and at that time they were polling personal approval ratings for the state leaders, each electorate's individual members and each electorate's federal member.  There was also a run of similar polling reported in the leadup to this poll.  It is not yet clear whether or when this data - probably much more informative than preferred-Premier polling - will be released.

State Polling Aggregate

I maintain an aggregate of all Tasmanian polling to try to smooth out the quirks caused by individual results.  There hasn't been a large-sample Tasmanian poll since the last EMRS, and the change in Prime Minister is a factor likely to shift state-level polling, so I've decided to give the new poll an unusually high weighting (for EMRS) of 50% of the sample.

This is the new aggregate:


While I don't think Labor would take the Liberals' seat in Franklin in the current EMRS poll, most polls for a while have been consistent with that seat falling, and the aggregate reflects this.  The key seat is, as usual, Lyons.  Preferences for "Other" voters at the last election in Lyons, whether they were voting for PUP or not, favoured the major parties heavily, which would put the Liberals over the line for three and greatly close the gap between Labor and the Greens.

For the readings above, I have Labor's Lyons seat returning to the Greens, though not by very much.  This should be treated with a lot of caution since the Greens do not have an incumbent in the seat, and so it's possible their vote in Lyons will not hold up as well relative to the rest of the state as my model suggests. But for the time being the aggregate has changed to 13-8-4, which the model says is very slightly more likely than the old value of 13-9-3.

NB: I was asked in a media interview about comparisons with poor preferred Premier results for past Labor leaders Lennon, Bartlett, and Giddings.  What I forgot to mention in answering that question was that prior to this term EMRS included the Greens leader in the comparison.  They no longer do so, so Bryan Green's result is in that sense worse than the bad results for those former Premiers.  However, that largely reflects the advantage to the incumbent leader at the time.

2 comments:

  1. Great analysis Kevin! Is 5% the smallest ever difference between Labor and the Greens in an EMRS poll?

    ReplyDelete
  2. No, it's not even the smallest difference of this term (it was 25-21 in the May 2014 poll). The smallest ever difference was one point (23-22) in May 2012.

    ReplyDelete

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.