Monday, July 29, 2013

ReachTEL (State): The North Remembers

Yesterday's article Tas Federal BaByLon Still Falling? covered the Sunday Examiner's large sample-size ReachTEL of Bass, Braddon and Lyons, three at-risk federal electorates.  The poll showed Labor trailing in all but within striking distance in Bass and Lyons.  This poll challenged a widespread view that the return of Kevin Rudd as Prime Minister had fixed things in the state and that Labor would now retain most or all of its seats here.

The Examiner also asked a state voting intention question - just primaries and nothing else.  Below are the results with undecided voters removed, with the June results of the Mercury statewide ReachTEL shown in gray, and the swing from the June poll shown as well.





 The general patterns here are as follows:

* Labor is up across the three electorates, by an average of five points.  However, this increase compares with June results that were so miserable that some of them could not be taken entirely seriously.

* The Greens are down slightly on average, and in the one seat where they are shown as up (Braddon) the increase is to a percentage that is just not credible at the moment.

* The Liberal vote is overall about the same.  However, for the first time it is Lyons which the poll results suggest as a possible four-seat result for the party.

*  The Others vote, which seemed unnaturally high in June, has crashed.

It's the same old story and has been for most of this government's term.  The economy of northern and rural Tasmania is struggling so it's back to polling numbers not seen since Gray and Groom we go, with not much mercy for the coalition partners, like the Accord days all over again.  The Liberals are easily winning three seats in all three electorates and would only have to get two apiece in Franklin and Denison to win the election outright.  The party would, of course, prefer to win by more than just one seat, and this poll again puts them within striking range of getting a four-seat result somewhere in case they don't get three in any southern electorate.  Labor's "recovery" from the June poll is not good for any extra seats on these figures.

While the Greens would hold Braddon according to this poll, their sample result for Braddon is clearly excessive and not much should be read into it.

New State Aggregate

All that remains is to aggregate this poll with my previous aggregate.  These aggregates are based on the three-party-preferred vote, assuming for modelling purposes that there won't be any significant fourth party runs until such time as some declare themselves and poll something decent.

For these purposes I'm weighting the current poll, in the seats it applies to, at 40% and the previous aggregate at 60.  It is tempting to assume that the swing patterns in the rest of the state are similar to those displayed in this poll, and on that basis to also adjust the other seats, but we have enough evidence of non-uniform swing behaviour in Tasmanian polling lately that I've decided to leave those seats largely as they are. However I have made one change, which is to deduct half a percent from the Liberals and add that vote to Labor in every seat, as a precaution against ReachTEL methods being perhaps a shade Liberal-leaning (as is apparently the case nationally). Here is the new aggregate:


I've had some feedback that the Liberal vote in Denison in this model is too high but until public data comes along showing it to be lower I'm not going to subjectively change it.

At the moment the aggregate is very heavily ReachTEL-centred, since the last EMRS is two and a half months old.  That is why Labor's average has gone down slightly in this aggregate despite the current ReachTEL being better for Labor than the previous one.  Anyway, this will change when the new EMRS comes out in a few weeks from now and is included.

The current aggregate shows that it's still possible in theory for the Greens to hold all of their seats despite a quite large swing against them.  However, while the Greens would hold at least three seats if the above result played out exactly, their position is fragile and vulnerable to random variations in every electorate but Denison. There is some case for changing my current projection of 14-8-3 but since it is not clearly wrong I have decided not to.

While the Liberal Party remains strongly on course to gain at least three seats and probably more, it remains very unclear at whose expense its gains will be.

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. 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.