Legislative Council Voting Patterns 2022-6

In the leadup to the 2026 Legislative Council elections for Huon and Rosevears (link TBA when I've written it) this article is my annual review of voting patterns on divisions in the upper house in the previous four years.  But before I get into it, I need to deal with some methods nerdery at the start.

Shy Division Losing

Some Labor MLCs aren't particularly fond of my findings, and they were especially defensive about a stat that in the small sample added to the mix last year Labor had voted with the Liberal Government 90% of the time.  (Fear not, in this year's sample it is 86%).  This even led to an attempt on election night live TV to shoot down (but not shag or marry) my methods in which it was falsely claimed that if there were thirty divisions on a single Bill I would include them all in my assessment.  Fortunately the incorrect claim has since been retracted.  

What in particular the Labor MLCs do not like goes to an unfortunate quirk of the LegCo's standing orders. When votes are called for on a motion they are initially taken on the voices.  The President or whoever is in the chair at the time declares a provisional result, eg "I think the ayes have it".  At this point someone can call for a division - but only if they are voting on the side that lost the call on the voices.  

In a case where the Government has no friends on a vote they might vote one way on the voices, but then not bother having that vote recorded to avoid embarrassment.  And in this case, while Labor voted the other way, there is nothing Labor can do here to cause a division such that them voting on the other side shows up in my figures.  This does sometimes happen, though no evidence that it happens often has been presented.  This is in contrast to federal parliament, where divisions can be forced on the losing side by the winners provided that there were multiple voices.  A notable example was the final vote on same-sex marriage.  

The argument for the Tasmanian provision is presumably that having divisions to confirm what you already know about the outcome of the motion wastes time, so the ability to call for a division is there for cases where the losing side wants to check if it really lost.  If saving time at the expense of being able to keep an accurate record of voting behaviour is the real reason then it's curious to see it in a house that doesn't even have time limits on speeches, and probably it should be changed.  That said even if it was changed, shy governments could always let motions they would lose 11-3 go through on the voices without saying anything, and quite a bit of that happens already with this government.

So yes it is true that last year's direct statistic that Labor and Liberal voted together 90% of the time in a sample of recorded divisions doesn't capture unrecorded divisions and there's a skew in the sorts of divisions that go unrecorded.  One would have to watch the video of every motion to try to get a record of these.  But cases of the government being shy losers don't change the relativity of all the other motions.  In the last year, Labor voted against the government in only two recorded divisions that I found.  Every other MLC voted against the government between five and twelve times except for Tania Rattray (three) ... and Rattray is the Leader for the Government!

The more usual methods stuff

The methods for this article are the same as in previous years.   I look at the last four years of data on a rolling basis.  Where a given Bill or other matter has multiple votes, I include the divisions that are different to each other, up to a maximum of ten (not thirty) divisions per Bill.   While there are very rare conscience vote cases where members of a major party can be found on both sides of a vote (none this year), in general I treat "Labor" and "Liberal" as a single actor, and treat a party as absent for that purpose in the case of a split vote (there were no splits on either side this year).  I exclude lone dissents (none this year after accountinf for pairing). 

Note that Labor gives up a vote on every issue because Craig Farrell is the President.  There were no casting votes this year, for comments on the impact of casting votes see last year's article.  

In this year's rolling sample there are 78 divisions, of which 14 are new.  For all the fuss about it the Macquarie Point stadium accounted for only two of these.  Others included greyhound racing, petrol, policing, Marinus, parliamentary salaries and one that much fun was poked at concerning the length of Council sitting breaks on Tuesdays.  

Agreement matrix and left-right sort

The chart below shows how often each pair of Legislative Council entities (an entity can be an independent MLC or a party) agrees with each other on the contested divisions in the sample.   For instance this year's table finds that the two MLCs up for election, Dean Harriss and Jo Palmer (Liberal) vote together 62% of the time.  

There are as usual some clusters of high agreement percentages.  Webb votes over 70% of the time with both O'Connor and Gaffney though the latter two don't vote with each other all that often.  The independents Hiscutt, Rattray, Armitage and Harriss all vote with each other fairly often.  Armitage fairly often votes with the majors and the majors often (and increasingly) vote with each other.  There's not enough data re Hiscutt after only twelve votes but there's a cluster of the majors, Rattray, Armitage and Harriss (though Harriss often doesn't vote with Labor) and this leaves Forrest (who has high agreement percentages with some MLCs in both the left and right cluster) and Thomas (who doesn't get outside the 40-60 band with anyone except Forrest.)  Clearly Forrest and Thomas are in the centre between the 3-MLC left cluster and everybody else.  

This year for the first time, Labor has a higher agreement percentage with the Liberals than every other MLC has with the Liberals, including independents often regarded as "conservative".  This is a sign of the extent to which independents vs majors is becoming an axis in chamber voting.  Dean Harriss, for instance, voted differently to the government seven times this year.

I should preface the following by noting that the Tasmanian Liberal government has long been remarkably moderate by national standards and that referring to it or those who agree with it as "right" aligned is purely relative - nothing in Tasmanian politics is currently all that right-wing by federal standards.  I have included a ratio score for agreement with the left (O'Connor, Webb, Gaffney) and right (Rattray, Harriss, Armitage, Labor, Liberal) groupings - green shading means more agreement with the former, blue the latter (Forrest and Thomas agreed slightly more likely with the right group but I have given them a different colour as obvious centre MLCs) - but this is open to a challenge as concerns the most prestigious item here - mirror mirror on the wall, who is the leftest of us all?

The only alignment change from 2025 is that Thomas is moved from centre-left to centre, largely because of having more data.  I have tentatively placed Hiscutt as centre-right but this is only off 12 divisions.

In my alignment score Cassy O'Connor has come up with a very slightly (and statistically insignificantly) lefter ratio than Webb in the current sample despite voting with the Liberals three times more often.  Here there's a very debatable quirk of my ratio system.  There were two motions this year - disallowing parliamentary pay rises and referring the government's greyhound racing ban to committee - where the sole Greens MLC O'Connor voted with the government but nobody else did.  When this happens, Webb is marked as agreeing with Rattray, Harriss, Armitage and Labor (giving her an average agreement score with the right cluster for that item of 0.8) and O'Connor is marked as agreeing only with the Liberals (giving her an average of 0.2) so the ratio method in effect concludes that voting with the government and against the rest of the chamber's right is a left-wing act (you conservative Meg Webb you).  Which doesn't seem right.

The funny thing here however is that one of those motions was greyhound racing, where the government was actually rushing to implement a policy that in Tasmania is generally viewed as left (just don't tell Mike Baird that), and the independents all at least wanted more scrutiny.  So the counter-argument here is that when one of the left MLCs agrees with the government in cases where none of the more "conservative" independents or Labor does, they're probably showing a strongly left desire to just get on with it without delay while the government is just being unusually radical, and so the table's treatment for that item is correct.  A different quirk can come up with MPs way to the right of the government - there aren't any at the moment but I think of things like the federal social media ban which was a classic example of the circle theory of politics, opposed by the Greens and by several of the more fringe right Senators.  

Overall the government continues to get its way pretty often.  Aside from the two cases mentioned above its only other loss (and Labor's sole loss for the year) was on the second reading of the Youth Justice Facility Development Bill 2025.  Labor, Liberal and Mike Gaffney voted for, everybody else voted against and it was lost 6-8.

Anyway I hope these findings are of interest ahead of this year's elections.  I am considering doing one for the lower house since the 10 election at the end of 2026.