Saturday, May 1, 2021

Tasmanian Election Day 2021

Yes the glorious day has arrived, we are finally there!  We are (checks watch), er, ten months early.  

Tonight I will be doing live coverage for The Mercury.  The link will be edited in to this article when available.  It may be paywalled but there is usually a cheap introductory subscription for non-subscribers.  My live blog for the Mercury will probably start somewhere in the 6-6:30 range and go until late night although I may have to stop it to file articles and do interviews eventually.  I will be based at the tally room.  I ask media outside of the Mercury not to contact me by phone or email between 5:30 and about 11 tonight; once I have finished the live coverage I should be available for other interviews (feel free to say hi in the tally room when I don't look too busy to arrange).   Scrutineers are very welcome to send me news and figures by phone or email.

Link to coverage:

I don't know if anyone is doing an exit poll (there was one in 2018, which was quite accurate.)

My plan is to post postcount threads overnight (between 1-4 am) for all five Assembly electorates and for any Legislative Council division that is still in doubt or of interest.  A wrapup thread (see 2018 example) may or may not follow tomorrow or in coming days.  Unlike 2018 I will be home tomorrow and available for interviews but not before 11 or between 2:30-4.  

My main guide page to this election is here, with links near the top to individual electorates and other articles.

For those seeking voting advice, I recommend to number all the boxes.  This may make your vote more powerful and it cannot harm your preferred candidates.  If you vote 1-5 (or 1-6) for a party and stop, your vote can play no role in determining which other parties are successful.  Check that you have not doubled or skipped any numbers, especially not between 1 and 5.

Concerning the result, I have gone on strike from my usual practice of publishing projections because of insufficient polling at this election.  There is a commentariat consensus that the result will be the status quo in every electorate, totalling 13 Liberal 9 Labor 2 Green 1 Independent (either Sue Hickey or Kristie Johnston).  In my best attempts to aggregate what polls exist I have seen nothing that overturns this, but it is a herded consensus that could be wrong in either direction, and it is hard to know what weight to place upon the only voting intentions poll seen in an extraordinary campaign.  If the government gains seats it will be a remarkable result given its age and what we know about federal drag.  Indeed, it will be the first case of a same-party government gaining seats when neither the state nor federal party were in their first term since Queensland 1972.  If the government loses its majority, even if it remains in office, the early election call will have been a demoralising bungle and a massive waste of political capital.  To see a Labor government I would have to assume the government falls short and Labor wants to take it despite saying they wouldn't and the crossbench prefers them.  If there is a minority government then it will have legitimacy issues because both parties promised to only govern in majority.  

A few things about the counting tonight:

1. There are about 40 dual booths that are Legislative Council and House of Assembly combined booths.  In these, the primary votes for the LegCo will be counted first.  I have my trusty LegCo projectrons all ready to go and the early LegCo booths may be useful in getting a handle on the state result, especially in Derwent.

2. There has been a large increase in prepoll voting which is likely to account for around 18% of all votes cast, up from just over 10%.  This is still very modest in comparison to other states.  There is My estimate is around 64% of all votes cast will be ordinary polling booth votes on the day.  The TEC intends to count some of the prepoll places tonight - typically the larger ones - and some others tomorrow.  Those being counted tomorrow are mostly outlying towns or adjacent-electorate prepoll places, but also include some of the dual booth prepolls such as New Norfolk.  I do not yet know the timing of the counts tomorrow.  The TEC will also count some of the postals.  All this means that the counts at the end of the night should be a little short of where they were in 2018 but probably not far short.  

3.  Usually most seats are known on the night at party level.  At each of the past two elections there were two seats that were in serious doubt at party level at the end of the night, though in 2014 there were a couple of others with small levels of doubt.  There are usually also a few that are unresolved at candidate level.  This year it is possible Clark will be ultra-messy with multiple unresolved seats at both levels, but maybe some people will win or lose easily and it will be a fizzer.  


I have been running Not-A-Polls in the sidebar to get readers' estimates of how many seats each party will win.  Thanks to all who've voted!   Unfortunately I started the Liberal one a long time before the others rather than running them all on the same time scale.  It's notable that the averages (currently running at Liberal 13.1 Labor 9.3 Greens 2.1 Clark INDs 1.15) add to slightly more than 25.  They might have done so by more had I canvassed the chances of an Independent winning outside Clark.   Overall though the reader pick (barring a Garland miracle) appears to be 13-9-2-1.  Readers of this site skew to the left on average and underestimated the Liberals by one seat on average in 2018 and were doing so by two seats in 2014 when I had to terminate the Not-A-Poll because of stacking.  If that trend is followed the Liberals could be on for a good night.

However, the end-of-election state of the Not-A-Poll in 2021 is not greatly different to 2018.  These graphs show the rolling five-day figures for the percentage of Not-A-Poll voters who expected a Liberal majority in each year. Firstly 2018:

Very few readers thought the Liberals would win a majority early in the 2018 campaign but as the campaign went on the share increased, peaking following polls on 24 Feb and 27 Feb that showed them to be in the box seat.  

Now 2021:

Ignore the dip around 3 April which is based on a very small sample of votes at that time.  Overall in the first few weeks around 70-80% of site voters thought the Liberals would retain a majority, but in the last two weeks this dropped off to around 50%.  That is still somewhat higher than the final figure in 2018 (when the Liberals only won by one seat but still won comfortably, given that a 6% swing would have been needed to change any result).  

I should note that Not-A-Poll allows multiple voting, so many people have voted at least three or four times as the maximum one-week IP block has expired.

Hope everyone enjoys the coverage tonight, even if some of you don't enjoy the result!  


  1. Hi Kevin,

    Re: Clark, has there ever been an electorate result where no candidate makes a quota on primaries? I could see that possibly happening in Clark today.

    1. Since the reduction to 25 seats this has happened four times: Bass 1998, Lyons 1998, Denison (now Clark) 2010, Lyons 2010. Both the Lyons cases nobody was even close. For the 1998 election having more incumbents than seats would have contributed to it.

  2. And to think, if the Labor Party didn't attempt to ban poker machines 3 years ago Rebecca White would have been reelected tonight.

    Interested if Craig Garland still an outside chance of getting up in Braddon?

    1. Alternative opinion, if the MOU with the pokie lobby hadn't been lied about and kept secret the ALP may well have won enough trust to win majority tonight.

    2. No chance for Garland, see my Braddon comments. Did pretty well though.

  3. As there’s still a possibility of no majority until Clark distribution of preferences, are we still in caretaker mode? I’m presuming no ministry swearing in etc. until then.

    1. Yes the new ministry hasn't been sworn in yet. Aside from the issue of Clark there is also the question of whether Roger Jaensch loses requiring a reshuffle of his positions. The new ministry is sworn in within a week of the return of the writ.