Hobart City Council Voting Patterns 2018-22

Advance Summary

1. Traditionally, on contested votes the Hobart City Council is loosely divided between "pro-development" councillors and councillors who stress environmental issues and/or the interests of impacted residents. 

2. This term of Council initially continued the pattern of the previous term in which councillors voted fairly distinctively and voting patterns were hard to firmly classify.

3. From around early 2020, however, the Council's voting on contested motions became far more factionalised, at the same time as such motions becoming less common.

4. As a result, while all councillors vote independently on particular motions, all councillors in this term can be classified as at least overall leaning towards the "pro-development" ("blue") mindset or its opponent ("green").  

5. Indeed, this term has seen some of the most polarised patterns in voting on contested motions in the last several terms of Council.

6. A possible ordering of councillors from "greenest" to "bluest" in this term is: Burnet, Harvey, Dutta, Reynolds, Fox, Sherlock, Sexton, Briscoe, Thomas, Denison (no longer on council), Coats, Behrakis, Zucco. 

7. Each of the "green" and "blue" clusters includes both more diehard members who are usually party-associated and also a more moderate sub-group.  

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

Hobart City is gearing up for an election campaign, with local government candidates to be announced next week and voting to occur by post from 3 October.  This will be the first election with compulsory voting and the first election since the state government reduced the number of preferences needed for a formal councillor vote to 5 to counter rampantly high informal voting levels.  Unfortunately it won't be the first election where nominating is far too easy, and as a result there will be too many candidates (about 33 are known already thanks to Darcy Murphy's tracking and others, including all the incumbents bar Sexton).  

In the leadup many candidates are characterising their political positions, using words like "independent", "progressive" and "teal".  (Teal has been such an overused colour on the campaign trail that I've changed my site to the opposite of teal until the election is over in the interests of chromatic balance).  For new candidates it's often hard to know if there's any truth in these labels; we can only judge them by their policies, party associations and who they align with.  More of this in my guide to Hobart Council which I am hoping to post on Sunday Monday.

But for existing councillors there is a wealth of data in the form of council voting patterns for those who want to see who (if anyone) is the real teal and who might be astro-tealing, teal-stealing, teal-washing or maybe even teal-trolling.  I've been monitoring voting patterns for Hobart Council all the way back to the 2005-7 term, my last two offerings being on this site for the 2011-4 term and the 2014-8 term.  In general I have found that most Hobart Councillors, while all maintaining a degree of independence from each other, can be put into two clusters, which I call the "green" cluster and the "blue" cluster.  Members of the green cluster aren't necessarily Greens, but they tend to agree with the Greens more often than not, and often but not always have present or past connections with the party.  Members of the blue side aren't necessarily Liberals, but they tend to agree with the known Liberals more often than not, and they often too have present or past connections with the party.  Councillors with Labor associations sometimes slot in on either side of this spectrum, or are sometimes in the middle.  Now and then a councillor doesn't align strongly with either group, and some councillors might normally align to a moderate degree with one group or the other but be more or less unaligned in a specific term, usually late in their career.

The green/blue division most often emerges on development issues, including contentious building approvals (where the blue side will tend to support approval but the green side will tend support  environmental, aesthetic or neighbouring residential concerns).  However, it also appeared in this term across a wide range of issues including COVID management, the Taste of Tasmania and internal council processes.  As a result even some councillors whose position on building and change of use approvals doesn't fall into the standard blue or green pattern may still tend to fall on one side of this line most of the time.  

Over the last few decades the Council usually had more blues than greens, but the blues struggled to capture leadership positions until Damon Thomas (2011-14 term) and Sue Hickey (2014-8 term) won the mayoralty with Ron Christie winning the Deputy position for both these terms.  Things changed at the 2018 election following Hickey's move to state parliament, Christie's brief but tumultuous stint as acting Mayor and a campaign dominated by the proposed kunanyi/Mt Wellington cable car.  Ex-Green Anna Reynolds won the mayoralty by a huge margin with endorsed Green Helen Burnet doing the same as Deputy. Christie lost his seat and the election returned a council that appeared to be very finely balanced.

Methods

The great majority of council motions are unanimous.  Often a proposed development doesn't actually bother anyone but council approval is required; these motions aren't politically interesting but it is worth noting there's a lot that councillors of different viewpoints do work together on productively. Indeed, there's some evidence they've been doing this more often, since the number of contested motions in my sample fell by nearly 100 in spite of the sample period being slightly longer.  (The COVID pandemic which knocked out a few scheduled meetings and stopped a lot of business/building processes temporarily would also be a part of this.)

My interest here is in the politically informative contested motions, especially those with more than one vote on both sides.  I went through all the Council minutes that have been posted for this term and recorded voting on every non-unanimous motion with a "voting record" showing.  (This can arise through divisions or through a small number of councillors asking that their dissent be recorded.)  Where a vote repeats the same pattern as an earlier vote on the same agenda item I ignore the duplicate unless it has a different winner (this mainly happens where there are ties with the voting lineup flipped - which is relevant to my assessment of which Councillors are most often on the winning and losing sides.)  If an agenda item occurs across multiple meetings (eg a resubmission of a rejected proposal) I treat the two items as distinct as frequently the conditions of the item will have changed.  

It took about 11 hours to key in all the data.  In all I used 397 divisions of which 318 had two or more votes on each side (only the latter are used in the agreement matrix and main assessment of voting patterns.)  There's been some tendency over time for the more polarised councillors to be the more common lone dissenters, but it's not all that reliable, and whether or not councillors can be bothered asking for their disagreement to be recorded seems to be a matter of attitude as much as politics.

For each pair of councillors I find all multiply contested votes where they both voted and find how often they agreed and disagreed with each other.  This results in an agreement matrix which I then sort into loose groups, calculate agreement ratios involving councillors from each group and then put in order.  This Council term saw one mid-term change in membership with Will Coats replacing Tanya Denison; as they were never on Council together I've come up with a best guess (81%) of how often they would have voted together if they were.

My sample for this article goes up to 29 August 22, and includes only full Council meetings, not committees.  A note that I don't have time to do these for every council though I am aiming to at least do a smaller scale version for Clarence council (possibly based on subsampling) before voting opens.  The manual data entry for this article is not guaranteed to be error-free, but any errors should be insignificant to the overall patterns.  

Participation in contested votes

The following is my record of how often each councillor voted on contested motions (including lone dissents) that are recorded in my sample (which as discussed above omits some duplicates).  Council meetings are held roughly every two weeks, but were less frequent for a time during 2020.  Councillors are often absent from contested votes because they are on leave, either travelling away from Hobart (often on official council business including "professional development") or with other commitments.  Councillors must also abstain on motions they have conflicts of interest on.  Conflict of interest is, mostly, quite broadly defined and can even include just having commented strongly on an issue before it came before Council.  However, absence (for all or part of a meeting) is by far the larger cause of not voting for most councillors. 

These figures are not the same as attendance figures (simple ratios of how many meetings a councillor was present at), which tend to be somewhat higher.  Raw attendance figures can be somewhat misleading because some councillors are more likely than others to attend only part of a meeting for whatever reason, and some are more likely to abstain through conflict of interest.  On the other hand, a councillor with a high attendance rate might just happen to miss the meeting at which there were 18 contested divisions.  For whatever reasons there were a lot more contested votes in the first 12 months of this term than in later 12 month periods (something which also occurred to a lesser extent in the previous term.) I give what I get, unofficially, as the rate of recorded meeting attendance up til the end of August 2022 in the final column.

There was an unusual case in this term where a snap no confidence motion in Lord Mayor Reynolds was moved and she was marked as abstaining with a note that her abstention was counted as a vote against (I have counted this as being present and a vote against for voting intention purposes) - I believe this is what happens when a councillor abstains without leaving the room (which is easier done when a motion is on notice).  In any case the motion failed irrespective of Reynolds' vote.  (It was supported by Marti Zucco, Jeff Briscoe, Tanya Denison and Simon Behrakis.)

Within this term Jeff Briscoe missed no meetings and only three contested motions.  At the other end of the scale, Peter Sexton (a doctor with a historically heavy load of other commitments) missed about a fifth of meetings and nearly a third of contested votes in my sample, which is similar to the previous term.  

Lone Dissents And Losing Votes

Some other things I look at include who casts the most lone dissents and who is most often on the losing side of motions.  The following table is ordered by who is most often on the losing side of multiply contested motions (those which have at least two votes on either side).  It also includes each councillor's rate of lone-dissenting (out of all motions they voted on) and how often each councillor is on the losing side of all motions.

Lone dissents were not a huge thing in this Council term.  Behrakis and Helen Burnet had sixteen each and Mike Dutta and Bill Harvey each had eleven.  Will Coats and Behrakis were each on the losing side of multiply contested motions more often than not, while Sexton, Thomas and Sherlock were least often on the losing side.  (Sexton was also the least often on the losing side in the previous term.)  

It's common for Lord Mayors to place themselves above the fray and try to cast fewer lone dissents and get voted down less often.  While Reynolds' stats have followed this pattern I'm not sure there's any deliberate intention behind it - rather I think that without her position changing, there has been more polarised voting from other councillors and hence more motions she has avoided being on the losing side of.  It's notable in this table that both at the top and the bottom of the table the rates of being on the losing side have risen - reflecting not just a lot of motions with a few of the more hardline councillors on one side, but also an increase in very close votes and especially ties.  My sample contained 44 ties  in this term compared to 23 in 2014-8.  

Agreement Matrix

The following matrix shows how often (as a percentage) each councillor votes with each other councillor, subject to them both being present and to there being at least two councillors on each side of the vote.  For instance, the matrix shows Burnet and Damon Thomas vote together in such contested motions 37% of the time. With apologies to colour-challenged readers, I've used dark green and dark blue to highlight agreement scores above 75% (a commonly used cutoff for funding clusters in this kind of work) and lighter green and blue to highlight scores that are close to that level, in the range 70-74%.  

A strong disclaimer here that "Green" means small-g green and not the party (perhaps I should find some other word for it; I'm open to suggestions but am not sure left/right is always accurate for council business either.)

The letters in brackets show Greens and Liberal party endorsements where the councillor remained, to my knowledge, a member of the party while on council.  Jax Fox was endorsed by the Greens for the 2018 election but quit the party late in the voting period and was then elected anyway.  Tanya Denison quit the Liberal Party some time after leaving council.  Some of the other councillors may be party members but were not endorsed by any party as such.  (The Liberal endorsement was low-key in an email to supporters.)

The table shows two very obvious groupings.  One is the cluster of agreement percentages mostly above 70% with a few in the 60s involving Burnet, Harvey, Dutta, Reynolds, Fox and Zelinda Sherlock.  The other is the cluster of percentages mostly above 70% (in cases way higher) with several in the 60s and involving Sexton, Briscoe, Damon Thomas, Denison, Coats, Behrakis and Zucco.

However these clusters can be divided into sub-clusters.  Within the "green" cluster the endorsed Greens Burnet and Harvey are unsurprisingly the least likely to vote with the "blues" on contested motions (Burnet's agreement percentages of 13% with Zucco and 14% with Behrakis are the lowest I've ever recorded).  Reynolds, Fox and Sherlock all vote with each other quite a lot, but there is only one agreement percentage above 75% between them and the Greens (that being Reynolds/Harvey).  All of these councillors will vote with the official Greens fairly often but also have differences with them (for instance Fox might prioritise affordable housing over usual green-side objections to a proposal.) Dutta doesn't really fit in either sub-cluster (for this reason I have classed him merely as "green" and not "strongly green").  Given that Dutta is Sherlock's father but they have quite different political stories it was always interesting to see how often they voted together.  Dutta votes more often with Sherlock than with anyone else but Sherlock votes slightly more often with Reynolds and Fox.  Dutta often votes with Burnet against housing approvals (indeed more so than Harvey does) but is less predictably "left" across other issues.

Within the blues cluster there is the diehard grouping of Denison, Coats, Behrakis and Zucco (especially the last three named).  Behrakis and Zucco voted together on 92% of multiply contested motions, the highest percentage I've ever recorded - so while Zucco is technically an independent he votes with a hardline Liberal more often than a randomly selected Liberal probably would.  Then there is the more moderate grouping of Briscoe, Thomas and Sexton.  Sexton only has 75+% agreement with Thomas and has fairly high agreement percentages with the moderate greens - while I've put Sexton and Briscoe as "Moderately Blue" there's a case for calling them "Blue Leaning" instead.  

What was interesting in compiling this chart was watching the patterns change over time.  After the first several months I was seeing a lot of contested votes but not a lot of clustering, Briscoe was voting more green than blue and things were looking similar to the last term or perhaps even less predictable.  This changed around the start of the pandemic, with a number of 6-6 (or similar accounting for absences) green-blue votes on pandemic management issues.  From that point on while contested votes have been less common than in the first year, voting on contested votes has been among the most polarised and obviously factionalised that I have seen on this Council.  That's not to say that 6-6 votes predominate; there are a lot of votes that are most of the Council on one side with the endorsed Greens and/or Dutta on the other for instance, and also a lot that are most of the Council on one side with Behrakis, Coats and Zucco on the other.  (The cable car approval was the signal example of the latter.)

In two dimensions

Something I've always done with these voting pattern charts for Council is a two-dimensional principal components analysis graph to see if it picks up anything different on trying to squash the voting patterns into two variables instead of one.  What I generally find is that the first variable that the graph picks up is the left/right split and often the second variable doesn't explain anything much and is just driven by a single councillor's odd voting pattern.

This is this year's chart of the PCA loadings, and it's a little bit interesting:

It does rather strongly capture that the blues are two different clusters, and that Zucco and the three Liberals have very similar voting patterns but Thomas, Briscoe and Sexton are different, and not just a milder version of the same thing.

What's teal got to do with it?

So having explored all this, what can we say about the use of teal, especially by the Hobart Independents group that includes councillors Reynolds, Dutta and Sherlock?  In fact, Reynolds was an early adopter of teal in her 2018 campaign, but in her case it seemed to signify something different to what it did in 2022 federally.  A person who uses teal in a Liberal seat is telling voters that yes they are conscious of being in a Liberal seat and empathise with Liberal values, but they're greener than a Liberal - a safe vote for anyone who doesn't have a Labor bone in their body and who doesn't trust the Greens.  For the most part, teal has been most successful in seats Labor had no hope of winning.

Reynolds' use of teal in 2018 was a different message.  She had been elected as a Green but had decided to strike out on her own (having only tended to vote with the other Greens in the 2014-8 term anyway).  The colour there seems more like a signal that one is still in touch with a Green voter base, but offering something different for those wanting to move beyond the Greens.  (In spite of Burnet's popularity, there will never be any shortage of those.)  As such it's not surprising to see candidates who are similar in approach to the Greens but more moderate seek to co-opt the colour.  I don't see it as something aimed at soft Liberals in a city that on a two-party basis voted 70+% for Labor anyway.  

Overall when it comes to "independents" pretty much anyone who is not openly endorsed by a party can claim to be one, but this article shows that all the independents on the current Council have clear tendencies to vote one way or the other.  I suspect, or in cases can tell, that this would also be true of many of the independents running from off Council.

Disclaimer

Any comment on this article that reaches either me or anyone related to me will be considered on the public record, especially if stated otherwise.  This is to prevent candidates from attempting to covertly influence my commentary.  

Disclosure

In an entirely unplanned only-in-Hobart coincidence much of the data entry for this article was powered by a very good coffee purchased from Councillor Fox!