EMRS: No Real Change In Second Rockliff Era Poll (plus comments re 35 seats)

EMRS (Tasmania): Liberal 41 (+2) Labor 31 (+1) Greens 13 (=) IND 14 (-1) others 1 (-2)

Poll suggests Liberals largest party with Greens and/or Independents holding balance of power if election "held now"

Independent vote is probably being overestimated

It seems like the last one wasn't long ago (and indeed the interval was shorter than normal) but an EMRS poll of Tasmanian voting intention taken from the 8th to the 11th of August has just been released.  After an uptick in the previous poll that could have been caused by federal election contamination, the collective "others" are down three points to 15% in this poll, the voting intentions for which are almost identical to the March poll, the last under Peter Gutwein's leadership.  

The news seems moderately good for the now Rockliff government.  They hold a 10% lead, Labor has still to break out of the very low 30s, and nobody except the government is within cooee of forming a majority.  The government's lead would almost certainly not result in a majority if the votes recorded in the poll were cast as such at an election, but history suggests that doesn't matter.  Governments with slender leads or even no lead at all at points through the term perked up as the election approached (or well before it with a big hand from pandemic management) and won majorities in 2006, 2018 and 2021.  It's necessary to go back to the pre-EMRS 1996 election for the last case where a government with any sort of real polling lead did not win a majority.  The bandwagons seen in 2006 and 2018 may reflect the tactics of voters who want to avoid minority government, but it can also be argued that they were caused by campaign factors in those two elections.   

The poll would also come as some relief to the government in that it's the first time since August 2020 that the government has gone up, so the narrowing of the lead this year so far looks like a comedown off pandemic highs rather than a terminal decline.  It's also a sign that voters are not yet judging the government harshly for a remarkable string of resignations, several of which have had burnout as a significant theme.

The government is, again, below a majority-winning level of support in this poll, but the next election is a very long way away, should it be held on time - an interesting aspect of that being that both federal and state elections will be held in May 2025 if neither goes early or late.  

The Preferred Premier indicator shows Jeremy Rockliff leading Rebecca White 47-35 compared to 47-34 in June.  I am mildly surprised that Rockliff's lead did not increase given that he has had more time to become established.  In the general context of preferred-leader polling a 12% lead for an incumbent is not significant, though it is not too bad when the incumbent is new.  

The Mercury's article quotes Paul Jamrozik of EMRS as saying the poll shows that "Tasmanians have remained steadfast in their support for the Liberal Government over Labor by a significant 10 percentage points."  However voters who prefer the Greens, and to a lesser extent those who say they prefer independents, would tend to prefer Labor.  A 10 point lead is not necessarily enough, as the Gray government found out in 1989 when it beat Labor by 12.2% but got the boot.  

The high Independent vote in EMRS polls remains an interpretation problem.  Independents polled 6.22% at the 2021 election but 69% of that independent vote occurred in the division of Clark, with most of the rest in Braddon and negligible support for the obscure independents who ran elsewhere.  It is entirely possible the support for independents in Clark will grow further as Kristie Johnston becomes established (but then again it could drop if Sue Hickey doesn't run again or has little impact if she does), but it is very hard to see how the independent vote would more than double statewide without strong independent candidates.  It might be thought that Tasmania would be susceptible to teal-style campaigning but it is difficult to see this really catching on in Braddon and Lyons at least, and the dynamic driving much of the support for teals (that they were the only candidates who could beat the Liberal Party) does not apply in a proportional system.  Also, the divided nature of the division of Franklin makes it seem hard for a community independent type to get traction there.  

I suspect therefore that many of those choosing the "independent" option are either wishing they had an independent to vote for or else parking their vote between elections.  

35 seats: recent developments

There's been significant news in the move towards restoring the House of Assembly to 35 seats with Premier Rockliff releasing advice from Electoral Commissioner Andrew Hawkey on the issue of whether the restoration should be the old system of five seats with seven members in each (5x7) or instead a system of seven seats with five members in each (7x5).  Hawkey's report examines the practicality of 7x5 in detail, finding in summary that it would be confusing, expensive and would be likely to require the drawing of a boundary through the Hobart Council area community of interest.  On this basis the government has stated its intention to introduce legislation for 5x7 by the end of the year.

This followed comments by the Tasmanian Chamber of Commerce and Industry which had made the novel claim that 5x7 would "not produce members who can most effectively represent their constituencies", raising the obvious question of where this critique has been hiding during 24 years of 5x5.  There is, in fact, no evidence that the current system creates problems with representation for particular areas (eg the preselection of major party candidates from different parts of an electorate remains common), and no evidence that 7x5 would do a better job of it than 5x7.  Indeed, for a candidate with a very local support base, it would be easier to win under 5x7 than 7x5 because the vote they would get from their local area is very similar as a percentage of a quota (the quota in raw vote terms is very slightly lower under 7x5) but they would have a larger area to draw a small level of surrounding support from and more chances to win a seat without reaching quota because there would be two more vacancies.

The TCCI also raised a concern that I have raised, the potential for an increase in informal voting under 5x7 as a result of voters having to number seven boxes without error, but (i) their claim that an increase in informal voting would occur was overconfident - it would be likely but not certain in the absence of an education campaign (ii) the problem is simply solved by a transitional savings provision to continue to accept votes with a clear 1-5 as valid pending reviews of how many such votes there are and how the rate of them declines over the next few elections.

Given the government's emphatic statements on the matter, I am now assuming we are going back to 5x7.  Personally I commend the government for its response and for not thus far being swayed by contributions that look to be motivated by a desire to clip the wings of Greens and independents.  Seven seats simply is a better district magnitude than five in terms of more accurate proportional representation.  

My estimate of this poll at an election under the 5x7 system is that the Liberals would win 16 seats, Labor 12, Greens probably 4-5 and independents 2-3.  (It is notable that despite recent controversy over his vaccine and other social media views, Craig Garland again polled strongly at the Braddon federal election despite being outpolled by a Jacqui Lambie Network candidate).  Under the 25-seat system the most likely distribution would be 12-10-2-1 with the Liberals dropping a seat in Clark to (probably) Labor, though this would depend on Labor finding a good candidate.

I may add further comments on any reactions to this poll, though the days when parties felt obliged to greet every new EMRS poll with a barrage of spin appear to be fading compared to the early days of my coverage on this site.  I am often asked how previous elections would have worked out under the 5x7 system; a reminder that my full article on the subject is available here with updates at the bottom for elections since it was written.