Monday, April 19, 2021

Legislative Council Voting Patterns 2017-21

Advance summary:

1. This article presents a revised analysis of voting patterns in the Legislative Council (the upper house of Tasmanian Parliament) based on contested divisions involving the current (and one retiring) MLCs in the last four years.

2. Although there is a degree of independence in all Legislative Council voting (outside of party blocks), the Council continues to have a fairly clearly defined "left" side consisting of the five Labor Party MLCs, and independents Mike Gaffney, Ruth Forrest, Rob Valentine and Meg Webb. 

3. The three Liberal MLCs and retiring independent Ivan Dean form a "right" cluster and independents Rosemary Armitage and Tania Rattray now clearly more often vote with that cluster than with the left.

4. The votes recorded in the last year have also seen greater fracturing of the left cluster, more divisions with both major parties on one side and a higher proportion of divisions where the government wins.  

5. A possible left-to-right sort of the Council is Valentine, Webb, Gaffney tied with Forrest, the five Labor MLCs in no particular order, Rattray, Armitage, Dean, and the three Liberal MLCs in no particular order.

6. Going into the 2020 elections, the left holds an absolute majority in the Legislative Council, normally meaning that the government needs the support of Labor or at least two left independents to win votes.  The left will remain in majority, the question being the size of that majority.

7. The left-right split is only rarely reflected perfectly on the floor of the Council at the moment.


Two Legislative Council elections will, controversially, go to the polls on May 1 alongside the Tasmanian House of Assembly election.  My guide pages for the elections are here:



House of Assembly main page

There is no election in Mersey, where incumbent Mike Gaffney has been re-elected unopposed.  He is the first to be returned unopposed for ten years.  There were four such cases between 2005 and 2011 after a 12-year gap since the previous case in 1993.  

In most years I do an article about Legislative Council voting patterns as a curtain-raiser to these elections, but in this case the time window in which I would normally have done it has been swamped by the early election.  Click here for the 2016-20 edition, and click the "parliamentary voting patterns" tab at the bottom for articles including all the previous editions (and some other pieces.)

In 2020 a left-wing independent retired and was replaced by a Liberal, and a right-wing independent was defeated by Labor.  This made the Council a party-majority house for the first time ever, but did not much affect its balance (nine left, six right or centre-right).  In 2021 we have a right-wing independent retiring and a Labor MLC under challenge, so no matter what happens the left will still have a majority.  However, if the Liberals can win both of these seats, they will increase the right's numbers to seven.  This would then mean they would only need one of the left independents to vote with them - as happens sometimes - to win votes.  On the other hand, if they fail to pick up Windermere their position could become more difficult.  

Labor President Craig Farrell has not to my knowledge had to vote since taking over from Jim Wilkinson in 2019.  In the one case where a vote was tied, it was in the committee stages where the President does not vote - and even if he had a casting votes, normal chairing conventions would have seen him vote with Labor anyway.


I've continued to use the same methods as in the 2020 article, but a few comments are in order.  Firstly I've continued to treat the five Labor MLCs as one entity, and the three Liberal MLCs as another. While in theory the Labor MLCs might at some stage differ from each other on a conscience vote, that hasn't happened yet.  However, during the marathon End-of-Life Choices (Voluntary Assisted Dying) Bill 2020 committee stages, there were six cases of the three Liberal MLCs splitting on particular amendments or clauses.  Leonie Hiscutt was the odd one out in most of these cases, with Jo Palmer and Jane Howlett the odd one out once each.  

In general I've simply extracted Labor and Liberal positions from each vote, but the cases where the Liberals exercised conscience votes on both sides are trickier to deal with.  Treating them as separate MPs who nearly always agree with each other would result in the high agreement percentages swamping the ratios I calculate, so I decided it was best to simply treat "Liberal" as absent from these votes.  Some people may wonder why I include conscience vote issues at all, but I do think they're predictive to some degree of broader political attitudes.  However I don't think six votes on a single issue is enough information to decide who is more to the right out of Hiscutt, Howlett and Palmer.  

I again use only the last four years of data and I only use recorded divisions with at least two votes (including pairs for absence) on either side. Where an issue generates multiple divisions on the same day, I ignore any of the divisions that are exact repeats of the earlier ones.  Where an issue generates identical divisions across different days, I've in the past included both.  But I've also decided now to not count such duplicates where the total number of different divisions on the one issue is ten or more.  (There was one such duplicate in the End-of-Life Choices case).  

I've added 31 contested divisions since the last report, ten of them being the End-Of-Life choices case.  Aside from that the number was down because of COVID, with a period from May to mid-September with very few contested votes.  The other big item was land planning laws (7 votes) and the rest were quite eclectic.  There were very few contested motions that were purely procedural (I think only one, though I find these are often informative anyway).

Agreement matrix and left-right sort

This chart shows how often the Legislative Councillors agree with each other on contested votes.  For instance, the chart shows that Armitage and Gaffney currently agree 43% of the time.  As usual I've highlighted agreement scores over three-quarters in red and dark blue, and scores close to that mark are highlighted in orange and pale blue.  Given that I've removed all the 100% (or nearly so) agreement scores for party MPs with their own party, the highest remaining agreement scores are 84% for Liberal/Dean, 80% for Valentine/Webb, and 77% for Valentine/Gaffney and Labor/Forrest.  The lowest are 17% for Webb/Liberal, 18% for Valentine/Liberal, 19% for Valentine/Dean  and 23% for Gaffney/Dean.

Valentine, Gaffney and Webb all have high agreement scores with each other.  Labor and Forrest agree often, and also fairly often agree with Valentine and in Forrest's case Gaffney.  This makes a left cluster of sorts, but it's somewhat messed up by the low agreement percentage for Labor-Webb especially, so it could also be treated as two related clusters, (Valentine, Gaffney, Webb) and (Labor, Forrest).

On the right the Liberals and Dean form a cluster.  Neither Armitage nor Rattray agree with anyone else more than 69% of the time, but they do agree with each other, the Liberals and Dean more than 60% of the time and with everyone else less than 50% of the time.  In previous years Armitage and Rattray have come out as centre or centre-right but over the past four years I now find them to be more clearly on the right of the Council on average, although not strictly part of the Liberal/Dean cluster.  This isn't just the influence of a particular social issue - even taking out the End-of-Life-Choices divisions, Armitage still voted with the Liberals on 16/21 contested divisions in the past year, for example.  Both Rattray and Armitage have moved around from right to centre to centre-right in my analyses over the years, and while they are clearly on the right of the Council in the last four years of my sample they're still well to the left of the Liberals.  

This year's "score" figure for alignment reflects the emergence of more moderate left and right sub-clusters (via fragmentation on the left and Armitage and Rattray agreeing more with the Liberals and less with the left) so I've used a different formula that is not comparable to past years.  The score figure for those on the left side is:

((Average agreement with Valentine, Gaffney, Webb) + 0.5* (Average agreement with Labor, Forrest))/((Average agreement with Liberal, Dean)+0.5*(Average agreement with Armitage, Rattray))

For the right side, it's the other way up.  

The "score" figure is hence an indicator of how strongly each MLC falls on one side of the left-right divide or the other, red/orange for left and light/dark blue for right.  The higher, the stronger the pattern. 

It may seem strange that Valentine agrees with Webb 80%, Valentine agrees with Labor 69% and Webb only agrees with Labor 50%.  The explanation for this is that Webb has not been there as long (two years) and in that time there has been an increase in "Laborial" divisions with both major parties on the same side and some combination of independents (typically not all of them) on the other.  In the past year these made up 16 of the 25 divisions on which the Liberals were united, increasing the Labor-Liberal agreement percentage over a four-year span from 23% to 31%.

The power of the block of five Labor MLCs is such that they lost only four divisions in this year, three of them being End-of-Life Choices amendments.  The government MLCs were together on the losing side of six divisions, and these were dominated by law and order issues (anti-protesting legislation, mandatory sentencing, an amendment in the repeal of begging laws, etc).  These divisions included the few cases of the 8-6 left-right pattern on the floor appearing in its purest form.  Overall despite not having the numbers upstairs the government has been reasonably successful in the past year in getting things passed, outside of law and order issues.  As politics goes back to normal post-COVID, the results of this year's election will have a big influence on how it goes with that going forward.

Ivan Dean's retirement marks the end (for now) of a tradition of staunchly conservative independents.  Decades ago these were probably the majority of the Legislative Council, but in the past several years right independents have tended to be more moderate than the Liberals, with only Dean and Armstrong keeping the Liberals company.  Armstrong was defeated and Dean has retired so we may not see such independents again.  This in my view reflects an increasing interest by the Liberal Party in contesting seats openly, and also Legislative Council elections becoming more high-profile politicised contests and less like local government.   

The unopposed re-election of Gaffney is interesting in light of these results.  His voting record suggests the Liberals would find life easier without him, so it seems surprising that they did not put up a token challenger in the hope that a statewide landslide coupled with voters voting the same way on both ballots could win them the seat.  However, perhaps running against Gaffney would have been badly received in the light of his massive work on the End-of-Life Choices bill, and/or perhaps the Liberals figured they just didn't need a well regarded independent who they probably couldn't have beaten anyway active in the campaign and clashing with their lower house and Windermere messaging. Or maybe they just couldn't find a candidate. 

In the past I have normally run a principal components analysis on these results, but it usually isn't informative (usually showing that the left-right axis explains everything that can be readily explained beyond issues where given MLCs vote individually and unpredictably).  Time is short amid this hectic election campaign so I am putting what I have here out now, and I may add more later.  

Added 23/4: Ivan Dean Comments

It is worth briefly noting an odd comment by Ivan Dean in the Examiner today:

"Of these seven Independent members there are some who closely align with Labor ideals and that is evidenced from their voting patterns.

In fact, one such member, to my knowledge, has never voted against the Labor party."

This is obviously false as my article shows, with no independent agreeing with Labor on more than 77% of contested motions in the last four years.  Indeed the independent who has most often agreed with any party is Dean himself (84% with Liberal!)  This wasn't the case in the past - my previous articles would always find at least one left-independent who sided with Labor more often than Dean sided with the Liberals - but the fragmentation of the left that I refer to above has driven the independent/ALP agreement rates down.

Dean also claims that if Labor wins both seats on May 1, they will be able to block legislation with the assistance of one independent.  But that's not true while Craig Farrell is President; the legislation would then pass 8-6.  (If Farrell were not President, it might be tied 7-7 giving the President a casting vote, and then blocked if the President followed chairing conventions.)

Dean has not done wonders for the image of the Council's independence in making these attacks in order to support his preferred independent candidate.


  1. Since the last elections in which the major parties jointly won a majority, has there yet been a vote with Labor and Liberal on one side and all the independents on the other?

    1. Not yet. A split with just one independent (Dean) voting with the majors, five against them and Forrest absent happened twice in the land use planning amendments.