Voting Patterns In The Tasmanian House Of Assembly (2014-2019)

Advance Summary

1. In the previous Tasmanian Lower House term, the most common voting pattern was Labor and the Greens voting together against the Liberal government. Cases of the Liberals and Greens voting together against Labor were very rare.

2. Between the last state election and September, the Government had significant defeats caused by renegade Liberal Speaker Sue Hickey sometimes voting against it, but Sue Hickey still voted with the Government more than 80% of the time.

3. Since ex-Labor Independent Madeleine Ogilvie rejoined the parliament, the government has not lost any votes, with Ogilvie almost always voting alongside it, and only voting against it so far on symbolic motions.

4. Since Ogilvie rejoined the parliament, Hickey's voting behaviour has become still more independent, to the point that she no longer strongly votes for or against any of the parties or Ogilvie.

5. Votes with the Liberals and Greens voting together against Labor have been significantly more common in this parliament than the previous two.  

--------------------------------------------------------------------------------------------------------------

Last week was a contentious week in Tasmania's lower house.  Tempers were frayed as the government used the support of ex-Labor independent Madeleine Ogilvie to pass mandatory sentencing and protection-from-protestors Bills through the Lower House.  These Bills are seen as having little or no chance of passing the Legislative Council irrespective of the outcome of its May elections, and attempts to weaponise these issues in past LegCo elections have failed dismally, but the government was crowing nonetheless.

Normally this site covers voting patterns in the Legislative Council.  (A 2019 update for that house was held over because of my workload on other elections and the lack of recent contested votes there; I expect to do an update prior to the 2020 poll.)  It's a very long time since I found House of Assembly voting patterns interesting enough to write about.  The last time was in the very early days on this site when I wrote The Compliant Coalitionists, which showed that a common perception among Greens supporters that the Greens were the real opposition to a major party cartel had simply never been empirically viable across the full range of issues parliament votes on.  The article also showed that the so-called "Laborial" voting pattern (Labor and the Liberals on one side, Greens on the other) became much less common when the Greens were supporting a Labor Government.  

Since the 2018 state election, voting patterns in the House of Assembly have suddenly become much more complex.  Previously there were only three moving parts and three possible voting patterns, excepting for very rare conscience votes and cases of crossing the floor.  However, interest was added when the Liberals' Sue Hickey, displeased that her business background and four years as Hobart Lord Mayor did not qualify her for a ministry, not only accepted help from Labor and the Greens to take the Speakership, but also started voting against her party on a range of issues.  Despite this, she remains a member of the party.  The situation became more complex still when one-term Labor MP Madeleine Ogilvie, defeated at the 2018 elections, returned to the House on a recount for Labor's departing Scott Bacon.  Ogilvie then chose to sit as an independent.

Mostly this article will explore the strange dynamics of the House with the influence firstly of Hickey and then of Hickey and Ogilvie.  However first it's time for a quick look at the Hodgman government's previous term.  

A brief explanation of my methods: I find contested votes by searching the Hansard Votes and Proceedings for the word "teller".  Where multiple votes with the same voting pattern occur on the same issue on the same day (eg amendments to a motion and then the final vote on the motion) I treat these as a single vote, to prevent amendment-rich debates from swamping the figures.  (Some documents are duplicated in the search so some votes may be included twice, but I made some effort to avoid this.) 

The Hodgman Government's First Term

The Hodgman Government governed in majority throughout its 2014-8 first term, holding 15 out of 25 seats.  Conscience votes on dying with dignity and same-sex marriage (the latter symbolic only) were the only votes to see the parties move out of block formation.  The following is the pie graph for which party served as the odd party out (yellow is the conscience votes):

Excluding the three conscience votes, the following was the pattern on contested votes (unanimous votes are ignored for this analysis):

* 57% of the time Labor and the Greens voted together against the Liberal government.

* 39% of the time the Liberals and Labor voted together against the Greens.

* Just 4% of the time the Liberals and Greens voted on one side against Labor.

The comparable figures for as much of the 2010-4 period as I covered before were 72%, 23% and 5%.  In Labor's last term of majority government, however, they were 8%, 41% and 51%.  And in the 1996-8 term of Liberal government with minimal Green support they were 34%, 24% and 43%.  

We can see from this that being on the same side of the fence makes a huge difference.  Labor and the Greens voted together and against the Liberals most often when in government together, then when in opposition together, less often when the Greens were giving minimal support to the Liberals, and hardly at all when Labor was in majority government.  Likewise, the Liberals and Greens voted together over half the time when both opposing a majority Labor government, almost half the time when the Greens were propping up a Liberal minority, but virtually never when Labor were in coalition with the Greens, and virtually never with the Liberals governing in majority.

(An amusing case of this rare event in the 2014-8 parliament was when the Liberals and Greens voted together to embarrass Rebecca White by suspending standing orders to make her give a speech explaining Labor's policy on poker machines. The speech did not receive rave reviews from its commissioners.)

The Messy Second Term: Hickey Phase

Prior to the return of Madeleine Ogilive the Hodgman government found itself in an unpleasant position.  Although it had won a majority and polled over half the popular vote, it had missed out by a whisker in the fight for Nic Street's seat in Franklin, and hence held only 13 of the 25 seats to 10 for Labor and 2 for the Greens.  

When Hickey started voting independently the government was placed in a quandary.  It could in theory expel her from the party (not for her voting behaviour since the Liberal Party only binds Cabinet members, but for some of her associated comments) but if it did this, it would no longer be able to say it was a majority government.  Furthermore it would make Hickey a victim, boosting her chance of re-election as an independent.  Therefore, the government couldn't really do anything about it, except put up with defeats on the floor while always blaming them on Labor and the Greens.  

Mostly, this was more embarrassing than anything else, because any government bill blocked by all of Labor, the Greens and Hickey would be unlikely to survive in the Legislative Council, where Labor holds four of the 15 seats and left-wing independents hold another five.  However, the balance of numbers allowed Labor, the Greens and Hickey to pass gender laws that were then passed into law by the Legislative Council.

A Speaker voting independently makes voting patterns more complex than someone else doing it, because the Speaker might vote or not vote on any given motion.  The perception is that the Speaker is in the Chair and only votes if there is a tie, but this doesn't apply during committee stages (consideration of amendments between the second and third reading votes) when the Speaker vacates the chair.  The Standing Orders also allow for the Speaker to vacate the Chair without warning or reason at any time and have the Deputy Speaker take over.  Labor's veteran Speaker Michael Polley did this to cast a vote against state-based same-sex marriage.  Hickey has now done this several times, but doesn't always do it.  This gives her the flexibility to either vote or abstain.

In a parliament where the four moving parts are Liberals (excluding Hickey), Labor, Greens and Hickey, there are ten different voting combinations possible.  There would be 2*2*3=12, but two of these would not create any division as the vote would be unanimous.  The following chart shows how often each occurred between the 2018 election and the return of Madeleine Ogilvie:

This table shows the various combinations according to whether Labor, the Greens and Hickey agreed with the Government's position (whether the Government was for or against the actual motion), the outcome for the Government, and the number of times that pattern occurred.  "Chair" means Hickey was in the chair and did not vote (if she gave a casting vote on a tied motion it is classed as yes or no.)  Three voting patterns never occurred: Hickey as the lone dissenter, the Greens and Hickey together against Labor and Liberal (though this one occurred after Ogilvie arrived), and the Liberals and Greens on one side with Hickey and Labor on the other.  

The (=)* voting pattern demands special explanation.  On a vote requiring a simple majority and with all seats filled, this pattern can't occur, because the vote on the floor is a tie and the Speaker must vote to break it.  However it did happen sometimes - firstly while Scott Bacon's seat was vacant (meaning the government won the vote 12-11 with Hickey not voting) and secondly on motions requiring a two-thirds majority (meaning no casting vote was required.)  In this case, the government won all these votes.

A majority government would normally win pretty much every vote, but in this case the government lost 13% of the time (and a larger percentage if counting individual divisions, because one of the issues where Hickey voted against the government spawned a very large number of amendments.)  Hickey broke rank on surgical terminations, gender law cases (on two different days), an adjournment motion, firearms laws and mandatory sentencing motions covering both sex offenders and frontline workers.  

Here is a contingency table showing on what percentage of votes each of the Liberals, Labor, Greens and Hickey voted together.  (For example to find the percentage for Labor/Hickey, read across the Labor line and down the Hickey column, or vice versa; it is 32%).  For Hickey this is confined to the votes she actually voted on, and in some rare cases those votes were cast based on convention (eg Hickey might vote for the second reading to allow more discussion but then against the third reading for policy reasons):

As concerns the three main parties, two-thirds of the time Labor and the Greens voted against the Liberals in this phase, while the remaining third of the time the Liberals voted with one of the other parties against the third.  "Laborial" votes were down from 39% in the previous parliament to 17%, and Liberal+Green vs Labor votes up from 4% to 17%.  It might seem that this is down to the small sample size of just 54 votes, but the change in voting patterns here is actually very statistically significant (chi-square 3x2, p=0.0002).

While Hickey crossed the floor on 18% of the votes she voted on (7/38) she still voted far more often with her own party on such votes than she did with Labor and the Greens.  

The Messy Second Term: Ogilvie Phase

The reappearance of Madeleine Ogilvie on the floor of the House was manna from heaven for a government that up til that point had barely seemed to take a trick.  If nothing else, dropping back from ten to nine seats was embarrassing for Labor, but it was soon to get worse, as the noises made by Ogilvie about stability have hardened into a pattern of being a key vote for the Liberals.  Ogilvie has presented some implausible arguments for voting the way she does, and has been willing to accept weak assurances from the Government or handball matters to the Legislative Council, when she could easily move amendments to refer matters to select committees or make sure matters she claimed to care about had been taken into account.

Why she is doing all this is a matter beyond the scope of this article (but not beyond the scope of comments to it, provided we can keep them legal).  But here is what it has so far looked like on the floor.  

Having Ogilvie as a moving part as well increases the number of possible voting patterns from 12 to 22 - and that's excluding conscience votes (none so far this term) and any case where Ogilvie might abstain or be absent (ditto).  One of those, the (=)* pattern, can only happen on a vote requiring a two-thirds majority, and hasn't appeared since Ogilvie was returned to parliament.  Remarkably, in just two and a half months of Parliament, 11 of the possible patterns have been seen already, in a sample of just 27 votes!  Here's the breakdown again:

Some patterns are especially notable:

O= and O-: these are patterns in which Ogilvie's key vote guarantees that the government will win the day.  In the case of O=, it's also possible Hickey would give a casting vote in the government's favour if she had to, but we don't know that.  In the case of O-, Ogilvie is definitely the difference between the motion passing and failing, and Ogilvie has cancelled out Hickey's vote allowing the government to win.  O- appeared in the mandatory sentencing and protection-from-protesters votes last week.

+: this is the pattern in which Hickey saves the day for the government.  However this has only happened once, and on a purely symbolic motion.

-: this is the pattern where the government loses because all of Labor, Green, Hickey and Ogilvie vote against a motion.  It hasn't happened yet and I won't be at all surprised if we get to the next election without seeing it (at least on substantive legislation as opposed to symbolic waffle motions.)

Now here's the contingency table since the arrival of Ogilvie:

This is a very small sample size, but some things can still be gleaned from this small sample.  Firstly, so far in her career as an Independent, Ogilvie has almost always (24/27) voted with the Liberal Government.  Moreover the three votes where she did not do this were symbolic waffle motions with no legislative consequence.  

Secondly, Labor has said it's making an effort to distance itself from voting with the Greens and to stop playing parliamentary games against the government.  However, it's too early to say for sure that this has worked based on such a small sample.  The percentage of such votes is lower than in the first half of the term, but the difference isn't statistically significant yet, either from the pre-Ogilvie phase of this term, or from the previous term.  In any case, the rate of Liberal-Green agreement may be returning to its very low rates of previous parliaments.  

Thirdly, Hickey in this sample votes like a true independent, as her agreement percentages with the Liberals, Labor, Greens and Ogilvie are all in the 30-70% range, without any strong clustering with anyone else.  If Hickey and Ogilvie were Legislative Councillors, we could say on this limited evidence that so far Ogilvie is a closet Liberal while Hickey is just a right-of-centre independent.  It's no wonder Greens leader Cassy O'Connor has taken to calling Ogilvie a "so-called independent".  (Ogilvie in the heat of the moment said she wouldn't vote with O'Connor again, which on the basis of the above patterns would make precious little difference if true, but would ensure the government could never lose a vote.)

Finally for completeness here is the contingency table for the whole 2018-22 term so far (in parliamentary terms, we're half-way through it, or very nearly so).  For Ogilvie and Hickey, only votes they took part in are included:

The decline in "Laborial" motions and increase in Liberal-Green vs Labor motions is still statistically significant compared to the previous term (p=0.001) but it will be interesting to see if this moderates as the term progresses.  Perhaps if Labor is firm in trying not to vote with the Greens we will see more "Laborial" motions and fewer Labor-Green motions, but the Liberals may continue trying to drive the Labor-Green statistic up with motions designed to wedge Labor into voting with the Greens.  (In the case of protection-from-protestors, the tactic seems to be to move a motion too defective for anyone but Ogilvie to support it.)

I intend to follow up this article at the end of 2020, and again at the end of the parliamentary term.