ACT Election 2024 Preview

2020 BASELINE: ALP 10 Lib 9 Green 6 (ALP-Green coalition government)

At election ALP 10 Lib 8 Green 6 FF1 (1 Liberal incumbent disendorsed and joined FF)

--

This is my preview page for the 2024 ACT election.  In the absence of any reputable polling whatsoever for the entire term (seriously!) there is not too much of use I can say about outcomes, beyond pointing to some possibilities.  What I can do is look at the nature of the contests based on the 2020 result and the history of ACT elections.  The ACT is of special interest to me because it uses a variant of Hare-Clark, the system also used in Tasmania.  I may get around to writing a special effective voting article but in case I don't here's one I prepared earlier.  Firstly, I've revised my view on something about the theory of ACT elections.

Well Yes There May Be A Bit Of Federal Drag In The ACT

Federal drag is the theory and fact that it is harder for state and territory governments to do well at elections when the same party is also in power federally, especially if the federal government is not that popular.  We saw this effect in spades in the recent NT election when the first of Labor's mainland dominos fell back with a drubbing that reduced Labor to 4 seats out of 25.  One could hardly blame the feds for all of that - though one could also blame them for more than some might like to.  (There was a lot of talk about how the Voice helped Labor hold up in the majority First Nations seat, the corollary of that is that it helped them get smashed everywhere else).  

Previously when I have looked at the federal drag effect in the very left-wing ACT I have not found anything to see - there just isn't a corellation between federal drag and government seat share change.  However, on a careful look at the history of ACT elections there seems to be something - you just need to ignore the swing from 1989 and 1992, and also looking at the balance of seats between the majors shows the effect up more strongly than whether a government gains or loses seat share, because the proportional size of the crossbench fluctuates a lot more than elsewhere.  The reason for ignoring 1989-1992 is that the 1989 ACT election saw a massive and largely once-off protest vote against self-government with the major parties managing only 37.7% of the vote between them.  Labor increased from 5 of 17 seats to 8 of 17 in 1992 in the face of federal drag, but this was mostly really about a protest vote in the 1989 election disappearing.  

From 1992 onwards, a kind of normality returned in terms of the majors getting most of the votes. In the early elections independents and ACT-specific parties were common crossbenchers, while lately the crossbenchers have been Greens.  In 1995, Labor were federally dragged near the end of the Keating government and lost two seats and government (they also lost a federal by-election in the ACT that year).  In 1998, the Carnell Liberal government was in its first term but was now federally dragged by the Howard government (also in its first term for what that's worth); the Carnell government broke even in seat terms and retained office.  In 2001, the now Humphries Liberal government was still federally dragged and lost office because Labor under Jon Stanhope gained two seats.  

In 2004 Labor, now in government opposite the feds, gained another seat and won the Territory's only ever majority.  In 2008 Labor was now in office federally and dropped two seats to the Liberals' one (all to the Greens).  In 2012 Labor, still in office federally and unpopular there, gained one seat but the Liberals gained two.  In 2016 the Coalition was back in government federally and Labor again became the largest party in the new 25-seat parliament.   In 2020, both major parties lost two seats to the Greens, the Liberals failing to make any inroads against a 19-year old government.  Overall, most ACT elections (2020 an exception) may show some seat-level pattern of federal drag, just not a very strong one.  

What the ACT doesn't seem to show (Labor dropping seats to the Greens in 2020 perhaps excepted) is a clear age of government effect.  Antony Green here notes the "forever government" effect and how rare this is these days.  In the ACT's case it helps that the ACT is so left-wing that it is very hard these days for the Liberals to win.

The Post-2020 Landscape And What It Means For 2024

I covered some of the anatomy of the 2020 result where the Greens won six seats in a previous article.  Since then there has been a redistribution with a substantial impact on two seats and a minimal impact on a third.  In brackets I note who the effective last-seat contest was in 2020, with the winner(s) before the "vs" and the losers afterwards).

Brindabella:  (Green 1 vs ALP 3) In 2020 Labor won 2.44 quotas, Liberals 2.31, Greens 0.65.  This may look like an easy win of the fifth seat for the Greens but it wasn't, because of what I call the "Ginninderra Effect" (old and slightly silly explainer piece here).  When a party with multiple quotas spreads its vote across a few candidates, it may defeat a single candidate who appears to be ahead on party total.  At a certain point of the distribution Labor had closed the gap to 2.59 quotas vs 0.71 quotas, but in candidate terms Johnathan Davis (Grn) was just 82 votes ahead of Taimus Werner-Gibbings. 

It's in general a reliable aspect of ACT elections that major parties will keep spreading their votes like this.  The redistribution was worse for Labor than for the Greens, but in a case like this the in-theory loss from the primary vote splits between three candidates.  As a result the Greens' notional lead over Labor comes out at about 0.011 quotas.  The Greens are especially vulnerable here to losing votes to minor parties, because any vote the Greens lose counts fully while any vote lost by Labor, all else equal, splits three ways.  So on paper the Greens' seat is precarious to say the least.  On the other hand, both majors have a long-term MP retiring, which doesn't do wonders for one's primary vote and preference flows in Hare-Clark.  

This is also the seat where the Greens, after very narrowly winning the seat last time, had a mid-term disaster which leaves them very vulnerable to fourth party competition.   On the other hand instead of having no incumbent they at least now have a part-term incumbent in the seat.

Ginninderra: (Green 1 and Lib 2 vs ALP 3) In 2020 Labor won 2.40 quotas, Liberals 1.60. Greens 0.75 and the Belco Party a very leaky 0.56 Q.  The Belco Party leaded Liberal preferences to be any shot but had fallen to 0.14 quotas behind the Liberals by the time they were excluded.  Overall Labor did best on preferences here, plus a small benefit from vote-spreading within their ticket, but it wasn't quite enough; at the key exclusion point Gordon Ramsay (ALP) on .789 quotas was excluded trailing Peter Cain (Lib) .807 Q and Jo Clay (GRN) .868 Q.  So Labor could gain the seat with a small swing off the Liberals or a relatively small one off the Greens (0.1 of a quota is 1.67%).  However, for the first 0.02 quotas Labor notionally gains, those votes are splitting two ways at candidate level.  

New entrants aside, this seat has become more complicated by the Liberals disendorsing their previous ticket-topper Elizabeth Kikkert, who is running for Family First.  They are also running only four candidates, which if they are excluded frees their fifth vote among those voters voting 1-5 and stopping to flow to Bill Stefaniak of Belco Party, Kikkert or anyone else perhaps still in the race (if indeed any of those are by that stage). Belco Party itself is only running three candidates, freeing its voters' preferences to flow to the Liberals (they ran five last time, which did not assist the Liberals who nearly lost the seat).  The seat has not been redistributed.

Kurrajong:  (Green 2 vs Lib 2) This was a major focus of my 2020 post-article.  Labor started with 2.28 quotas, Liberals 1.66, Greens 1.38 but the Greens overcame a 2324 primary vote gap to win their second seat by 470 (.056 quotas), the only time the Greens ever won two seats in a five-seat Hare-Clark division in either Tasmania or the ACT.  The win came primarily off Labor preferences and preferences from micro-parties, especially the 0.30 quotas polled by Canberra Progressives.  The Liberal campaign in this seat in 2020 was a disaster and they should be helped by the fact that their sole incumbent is now the party leader, which tends to provide a small boost.  On the other hand, the redistribution in this seat has been highly adverse to the Liberals, docking them 0.075 quotas and adding 0.05 to the Greens per Ben Raue's estimates.  (The draft redistribution was adopted with no changes.)  Notionally a small amount of the Liberal loss splits two ways but still they are starting something like 0.17Q behind (2.8%). 

Murrumbidgee: (Green 1 vs IND) Not much excitement here in 2020 with Labor 2.16 quotas Liberal 2.13 Green 0.7 with such challenge as there was to the 2-2-1 outcome coming from Fiona Carrick (IND) on 0.42 quotas.  After preferences the final margin was hefty, 0.345 quotas (5.75%) to the Greens.  The redistribution moves just over 0.02 Q from Labor to the Liberals but has little impact on the Greens.  The Liberals lose the leader bonus in this seat as Jeremy Hanson is no longer leader.  

Yerrabi: (Greens 1 vs Lib 3) Liberals 2.44 quotas Labor 2.05 Greens 0.61.  A seat of very little interest in the 2020 count, this eventually finished with the Greens' Andrew Braddock crossing the line 0.26 quotas clear of the second of two Liberals below quota.  But because there were two Liberals below quota, that figure underestimates how much Braddock won by, a Green to Liberal swing more like 0.17 Q (about 2.8%) rather than 0.13 Q would be needed to switch the seat, so about twice as "safe" as Kurrajong.  The seat was not redistributed.  

The Liberals lose their leader bonus in this seat, as Alastair Coe was leader in 2020 but is not anymore (indeed he quit the parliament mid-term).    

One thing that is notable about 2020 is that Labor achieved two quotas clean in every electorate, but were unlucky in not turning an excess of .34 quotas per seat average into any more seats.  This means that knocking Labor below ten seats requires significant damage.  In their weakest seat, Murrumbidgee, a swing of .16 quotas (2.7%) is needed to knock Labor below two quotas.  But even then, quota totals like 1.8 and 1.9 are almost invariably enough to win two seats in the ACT given all the ballot paper clutter parties polling fractions of a quota.  So there needs to be a pretty heavy swing against Labor from the 2020 result for Labor to fall below ten.  

This suggests that if the ACT's nearest equivalent of "federal drag" works in this election, the most likely way it works is the Liberals improving on their 2020 disaster and making slight gains, with Yerrabi and Kurrajong their nearest misses last time.  However this is complicated by ...

Fourth Parties and Independents (Or Not)

I'll start with this: the self-styled "Independents for Canberra" are not independents.  They are a registered political party that screens candidates for alignment with a set of principles and that has extensive party policies.  In the event of any members of this party getting elected and wanting to run again under the same banner they will be dependent on the party apparatus for re-endorsement and that will be influenced by what the party makes of whether they have adhereed to party principles.  In theory, elected I4C MLAs will be free to vote however they like and this is supposed to be the differentiation point with party politics. However, formalised parties that have claimed this have collapsed over differences in direction once elected before (eg Jacqui Lambie Network Tasmania, though JLN was much flakier about its expectations of candidates).  Also the free vote is subject to being assessed to have voted in line with party principles; anything else could lead to disendorsement for a future election, so how free is it really?  It's really Labor, with its understanding that crossing the floor even once can be career death (but isn't always) that is the odd one out here; it's quite common for minor parties to have some kind of theoretically free vote.     

The ACT's unfair electoral architecture encourages indies to form registered parties in order to have access to a column other than ungrouped, but Independents for Canberra are not a token party as a vehicle for ballot column access in the way "Fiona Carrick Independent" is such a token party vehicle for Carrick, who effectively is an independent.  They have the "independent" brand while also having the safety in numbers aspect of running group branding and shared policies to connect multiple candidates across electorates.  To me, this is trying to have your cake and eat it too.  

There are, in total, nine party groups besides Labor, Liberal and Greens contesting the election.  These are:

* Independents for Canberra, "Family First" (I cannot stand this party name so it can have scare quotes at its first mention too) and Animal Justice everywhere, though AJP are in the ungrouped column in Yerrabi because they only have one candidate there.

* First Nation Party in Brindabella, Kurrajong and Yerrabi

* Belco Party in Ginninderra and Yerrabi

* Labour DLP (seriously, this is still a thing? In the ACT? In 2024?) in Ginninderra and Kurrajong

* Libertarians in Ginninderra

* Strong Independents (vehicle for Peter Strong) in Kurrajong and Fiona Carrick Independents (vehicle for Carrick) in Murrumbidgee

* There are also ungrouped independent or unaligned candidates in every seat, none likely to be a seat threat this time around, and some of them serial candidates. (An unaligned candidate is a shy independent who declines to have "independent" on the ballot paper.)  

With the number of parties registering I was much fearing a ballot paper design disaster here but it could have been worse, with no seat having more than ten ballot columns.  All the same I think that next election the number of members required to register a party in the ACT should be increased.

Unusually, the ACT publishes candidate statements but the Greens and Canberra Liberals among others have declined to provide them.  

Independents for Canberra have impressively managed to field 20 candidates, including full slates in Murrumbidgee and Kurrajong.  Their lead candidate in the latter Thomas Emerson describes himself as the leader of Independents for Canberra (having a leader being another sign that I4C is a party) and was until May a staffer for David Pocock.  Emerson's name was easily the most mentioned in dispatches as a likely fourth party winner when I asked informed followers for their impressions on Twitter last night, several predicting he would win and many of those suggesting he would be the only one, or at least the only one from his ticket.  I have also however seen some commenters suggesting they will not win anything at all, one describing them as a "hot mess".  

One winning method here is obvious: a teal-style candidate who polls even as little as 0.4 quotas (6.7%) in Kurrajong rips votes from Labor and the Greens, tips out the second Green and beats the Liberals easily on Green preferences - unless, that is, the Greens are very good at spreading their votes between two candidates.  (Which is hard as this is their leader Shane Rattenbury's seat and he's very popular.)  That sort of vote is a fraction of what Pocock polled. Even though the strategic logic of voting for Pocock clearly doesn't apply in the ACT election, the candidate's close links to Pocock cannot hurt.  A Pocock-backed independent even won a single seat race in the Northern Territory, after all!

I4C appear to have vetted well thus far, with relatively few signs of the candidate implosions that often blight fourth parties.  The nearest thing so far has been Brindabella candidate Vanessa Picker saying that she would not support retaining Andrew Barr personally as Chief Minister.  This was in contrast to the party line that "We will follow our shared principles in seeking agreement on whom to support as chief minister".

In looking at I4C chances elsewhere it is worth being a little cautious about how strongly their vote will flow within the ticket; high primaries on the night will not necessarily convert to seats if the vote is not well concentrated with lead candidates (a problem that has ailed the Belco Party in the past, and also fourth parties in Tasmania).  But anywhere one looks at the 2020 results there are juicy partial quotas floating around to suggest that if I4C is even a modest success there will be seats it can win.  Absent of polling I have no robust idea how strong its chances are or which I4C candidates should be taken especially seriously.  However they are running the biggest fourth party campaign and I would not discount that they could win multiple seats.  But neither do I take it for granted that they will get anything at all.

Carrick is probably the next most talked about fourth party chance, though she has a pretty serious run as a 2020 baseline and has to improve on that or have the Greens go backwards quite a lot to win.  The Ginninderra situation with the departure of Kikkert potentially opening the door to the Belco Party's Bill Stefaniak or even Kikkert as an FF candidate (especially with the Liberals running only four candidates) is also interesting.  There isn't much buzz about the other minor parties.  

I have however had some reports that Belco are not that visible on the ground in Ginninderra and just seem to be going through the motions, while others think Stefaniak is a serious contender.  An interesting I4C lead candidate here is Mark Richardson, a serving police officer who is very high profile, eg after calling car hoons a "sub-species of the human race".  That pits him against Belco which historically is the pro-Summernats party.  

A common comment (a good indictment of it all by Jasper Lindell here) has been that the Canberra Liberals are still messy - they never seem able to reconcile fights between conservatives and moderates or to avoid having candidate disasters during the election.  In Ginninderra not only did they disendorse Kikkert - citing alleged disclosure irregularities and workplace culture problems - but also Darren Roberts, who is still on the ballot, ran into Facebook trouble.  A reader has also reminded me that the federal Liberals have a curious approach to making federal drag work for their team-mates, and in this term have continued (from opposition) to move a range of motions seeking to exploit the ACT's status as a non-state by overriding its legislation.

The Greens may have bitten off more than they could chew with their spectacular six-seat result in 2020, and now they have to chew like hell to keep them all or nearly so.  On paper three of their seats are highly marginal even ignoring the fourth-party threat; two of the others are somewhat loseable despite the substantial margins they were won by in 2020.   It should be noted that the Greens run corflute-free campaigns in the ACT, which obviously did them no harm whatsoever last time.  It is worth bearing in mind that in Kurrajong especially one party that would have deflated the Greens' primary last time (Canberra Progressives) is not running, and this could boost the Green vote in any competition with I4C or others.  

There has been a lot of attention on ballot order.  My experience is that ballot position in Hare-Clark is not very important; the places that use Hare-Clark are electorally literate and donkey voting tends to be low (it's not as much fun across the page.)  There is a case in Bass, Tasmania 2014 where it is known that the donkey vote in Bass (1-2-3 etc across the page in all forms combined) was at most 0.24% but probably below 0.2%, and the Australian Christians candidate polled 1.05% despite drawing pole position.  I just don't think that it matters.  

It is very hard, even with no polling, to see the Liberals getting anywhere near a majority, or even a combined majority for candidates who are clearly of the right.  It seems unlikely Labor would get one either especially given federal drag- there is in theory a path but they would need some swing to them and the Greens and I4C to poll poorly.  The big question here is how dissatisfied with the big three are voters really.  We have seen claims in at least the last few elections that voters were ready to pull the pin on them, but those claims came up short as fourth-party votes scattered among the ACT's riffraff of backyard parties such that none of those could win as votes exhausted.  This time seems a much more serious attack.  Assuming there is no majority, which of the following will we see:

* another Labor/Liberal/Greens shutout as has occurred every election from 2004?

* one or more non-big-three candidates elected, but the Greens retain sole balance of power?

* a Tasmania-style rainbow result where no party holds the sole balance of power and the largest major party has multiple paths to government?

* something even messier?

I should note that in Tasmania, the result was fuelled by the return to 35 seats (7 per division), but it would to some extent have happened under the 25-seat system anyway.  Having 5 seats per division makes life harder for fourth party challengers, but not impossible.  

Anyway I hope we get some actual polling in the next few weeks; going a whole election cycle without it isn't good enough.  I may add some notes to this article later, failing which I will be here for live and postcount coverage (but this year split into two threads); here's the 2020 thread as a taster for what you will get.  

Election guides

Tally Room

ABC