Tuesday, March 12, 2024

uComms: Labor Just 23: How Much Stock Should We Put In This?

This article is part of my 2024 Tasmanian election coverage - link to main page including links to electorate guides and effective voting advice.


uComms (Australia Institute) Liberal 37.1 Labor 23 Green 13.7 JLN 8.5 IND 12.8 others 5.0
Seat estimate if poll was accurate Lib 14 ALP 10 Green 4 JLN 2-3 IND 4-5
Poll should be treated with caution.

Today saw the the release of the third Tasmanian campaign poll by an established and identified pollster, this one being a uComms for the left-wing Australia Institute.

From the outset I should note some usual cautions.  uComms polls by automated phone polling (formerly all robopolling, lately a mix of SMS and voice robopolling).  The poll employs very primitive weighting (age, gender and location only, with no attempt to weight by any indicator of political engagement such as education).  At the 2021 election an Australia Institute uComms poll which I disputed at the time (What's This Then?  Commissioned Poll Claims Liberals In Trouble) was hopelessly inaccurate, underestimating the Liberals by over 7% and overestimating Labor by nearly 4 and independents by nearly 5.  There was never any attempt to explain why this poll got it so wrong.  

Many of the other cautions listed in the 2021 article still apply, though these uComms polls are using forced choice on voting intention rather than the previous option of allowing a respondent to be undecided on the first attempt then forcing on the second.  On the other hand, uComms just scored as close to a bullseye as seat polling gets with their poll of the Dunkley by-election.  

While this poll is more recent than the others going about (it was taken wholly on Mar 4-5 with a decent sample size of 1177) I would, for the above reasons, treat this poll with more caution than the fairly similar poll by EMRS.  And I suspect both polls could have the independent vote a bit high.  On my first of two campaign visits to Bass this weekend, the independents had almost zero corflute presence.

One more caution for this poll regards the Lambie Network.  There is no indication in the poll whether JLN was on the readout everywhere, so I am taking it that they were.   Based on Senate results from 2022, any poll that has JLN on the readout everywhere will overestimate their vote share by up to 14% of that share, which means the 8.5% of JLN could be really 7.3%, their weakest result this year.  That said the overestimate could be less as a fair proportion of Clark voters may have been aware JLN aren't on the ballot in Clark.  

But the really bad result in this poll, if this poll is anywhere near accurate, is Labor on just 23%.  This is lower than some numbers polled by the Greens in the 2010 election, and only barely higher than what the Greens in fact got.   I don't believe Labor is doing that badly, but to be doing badly enough that that number can even appear in a poll is a worry enough.  

In seat terms, using the EMRS breakdowns as a baseline, I get this poll at something like 14 Liberal, 10 Labor (some of those only just), 4 Green (outside chance of one or two more), 2-3 JLN and 4-5 independents (with a small chance of just 3).  But as with the Redbridge poll the combined major party vote is getting so low - it was 65 in EMRS, just 62 in Redbridge and now even lower at 60.1.  Can such things be true, or are some of these polls getting too many engaged voters in their samples?

I am intending to prepare an aggregate of all the recent polls but I expect it to show something broadly similar to the EMRS poll.  Polls continue to show the Liberals in a position to be easily the largest party and with multiple paths to retaining power among a mess of potential crossbenchers who in most cases have not indicated any clear preference.  Even the Greens have given Labor a free hit with leader Rosalie Woodruff saying she couldn't rule out working with the Liberals.  Labor has, however, to a large degree brought this on itself with its own disinterest in forming government via any kind of deal.  If voters are looking for a prominent candidate who is committed to throwing out the Liberals and to doing whatever deals can be reasonably done to have them thrown out, that voter will struggle.  

Choose Your Fighter!

This poll also included some fun leadership polls with the following unusual question form:

"If there is a hung parliament where no one party has a majority, which senior Liberal parliamentarian
do you believe would negotiate with the crossbench most effectively?"

Results are Jeremy Rockliff 39.4% Michael Ferguson 7.3% Eric Abetz 11.1% Don't know 42.3%.  Individually, over half of voters for each of Labor, Greens, JLN and Others were in the don't know camp.  The inclusion of Abetz is interesting as while he brings the gravitas of 28 years in the Senate he is not even a state parliamentarian at all yet.  The fact that he beats Ferguson including among supporters of his own party is an embarrasing number for the latter.  

Then ditto for Labor, and results are Rebecca White 31.2% Anita Dow 7.6% Dean Winter 19.4% and the great negotiator Don't Know was again in the lead with 41.8%.  This one is a bit odd too in that Dow has served as interim Leader and is now Deputy Leader but is not generally seen as a likely leadership option; perhaps Josh Willie would be the most likely third name to come up.  Anyway given that this is a sort of preferred leader poll albeit wierdly worded, the fairly high number for Winter is a signal of some discontent among voters generally with White's leadership.  This is as usual more so from voters who would not vote Labor as from those who would.  

A few more interesting snippets from the poll: those voters still voting Liberal overwhelmingly think Tasmania is heading in the right direction (75.1-9.5) but nobody else much does, and a plurality of Liberal voters think there will be a Liberal majority government (49.0%) compared to 4.2% of anybody else.  (This is in a choice between Liberal majority, Labor majority and hung parliament).

There were more questions in this poll yet to be released, including logging in "peace deal" forests, limiting rent increases to inflation, banning donations from gambling, property and salmon interests, a pony poll about having a real Integrity Commission, and a question about salmon farming that only the Australia Institute thinks is salient to voters.  I may comment on these later as they appear.  


  1. Can we expect the vote breakdown by electorate to be released?

    1. Doesn't seem likely at this stage. Would be interesting.

  2. Any one who takes a poll and then publish the results should be open as to the methodology. Including same size method used nature of any re weighting done.

    1. uComms are members of the Australian Polling Council but there are still more details that I would like to see than are in their APC disclosure statement.

  3. Well I guess we're about to find out whether "corflute presence" matters very much, in an era when so much campaigning is over the interwebs.

    1. There have been some papers from the US that have tried to find evidence of having more signs up increasing the vote for a candidate, and haven't found it. Also I believe ACT Greens have had success with no-corflute strategies. Nonetheless my general experience in Tas is that if a candidate has almost no signs up they're going to lose.


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.