Northern Territory Election 2024: Prospects and Polling

Welcome to my coverage of the Northern Territory 2024 election, which has snuck up on us all with just four weeks to go til polling day as I write.  NT politics often attracts derision among election-watchers for its tiny/barely inhabited electorates, crazy seat swings, frequent MP defections and elementary ethics fails.  Despite that though, I reject the view out there that NT elections deserve no more attention than a medium-large council.  NT elections are somewhat like state elections in their own unusual way, and are informative.  This one is something of a prelude to Queensland as the first chance for one of Labor's seven state and territory dominos to fall on PM Albanese's watch ... but will it, and how heavily if so?  Beyond this article, my coverage of NT 2024 will include a live article and a post-count piece (a la this, but they will be separate articles this year) and there may be prospects updates if there is anything to see.  

General properties of NT elections

The history of NT elections since self-determination splits neatly into two halves, 1974-1997 during which the CLP frequently changed Chief Ministers but invariably won, and 2001 onwards which, starting from Labor's first win under Clare Martin, has been a rollercoaster.  The 2020 election with a 2PP swing of 3.9% against the then Gunner Labor Government was in fact the most placid this century, with the five before it having swings (ALP) of 6.0, 11.1, -9.2, -5.1 and 13.3.  

The Northern Territory is affected (see bottom of article here), much as the states are, by what I call "federal drag" - all else being equal, incumbent governments are more likely to be whacked when the same party is in power federally.  2020 was one case where the swing went to the side in power federally, but that was off a ridiculously low base and still an easy win for Labor anyway.  

Because NT electorates have so few voters, many voters know their MLA personally and so NT elections have much larger personal vote effects than state and federal elections.  This has been a received wisdom in the psephology of the NT, but I hadn't seen anyone look at it for a long time so I checked it out myself.  I have found that on average since 1987 "sophomore" MPs (those defending their seat after winning it as a vacancy at the previous election) outperform their party's Territory swing by a mean 2.3%, median 1.4%.  "Double sophomores" (MPs who won their seat from the other side's incumbent at the previous election) outperform their party's swing by a mean of 3.2%, median 4.4%.  (The mean would be over 4% too but for the sample including Bess Price who lost the remote seat of Stuart, now Gwoja, at her first defence in 2016 with a 30.9% 2PP swing).  Vacant seats see the incumbent party underperform by mean 4.4%, median 2.2%.  Even when a sitting member has served only part of the term after retaining it for their party in by-election, that's still good for a deduction of mean 2.5%, median 1.4%.  Because of the wide variation in individual swings, causing the means and medians to be quite different, these figures give only a rough idea of the size of personal vote effects but they are clearly larger than elsewhere, where for instance "sophomore" and vacancy effects are worth about a point.  

NT elections also have high variations in swings between different seats in the same election.  Even in the sedate era of CLP dominance this was becoming a thing, with the 1997 election featuring an overall swing of 1.6% to CLP but also seeing double-digit swings in both directions in individual seats.  In the last four elections from 2008 on, the standard deviations in 2PP seat swings (albeit with some assistance from a one-election switch to optional preferences) have been 4.7%, 5.8%, 8.2% and 6.4%.  In compulsory voting states and federally the standard deviation is often in the 3-4.5% range. A factor here is that margin estimates for redistributed seats in the NT are unreliable.  

This particular election

The 2020 election was something of a return to whatever normal is for the Territory after the CLP was smashed down to two seats in 2016 following four years of absolute insanity (there's a book about it called Crocs In The Cabinet).   Labor won 14 seats, the CLP 8 and others 3 (independents 2 and the much-hyped Territory Alliance just scraped 1).  Labor won the 2PP 53.3-46.7 but were weighed down by a poor local economy in an election they would have won by more with the aid of COVID and federal drag.  COVID will not assist them anymore and the government is now eight years old and federally dragged itself, and all else being equal will lose seats, and perhaps several of them.  

Early in the term Labor historically won the CLP seat of Daly in a by-election.  They also retained Arafura with a massive swing to them after the death of the incumbent.  But things went pearshape for Labor late in 2023 after Michael Gunner's replacement Natasha Fyles came under pressure over a series of mining related conflicts of interest and eventually resigned.  Eva Lawler has been left to clean up the mess (and is widely considered to be making a good effort at that) but has not been helped by the continuing Alice Springs crime crisis that the government has had to fight with extra budgeted spending and on and off curfews.  

As I noted when she went Fyles is a rare case of a head of government who never faced an election; in three state-level cases of this the government rolled their second Premier for the term then lost the next election (two of those badly).  There is one prior case of a Territory government surviving under its third leader of the term, but that was when the second leader defected rather than resigning in disgrace.

Numbers go into the election at 14 Labor, 7 Liberal and 4 independents, but one independent is retiring at the election.  Counter-acting Labor's win of Daly, during the term Mark Turner (Blain) was for a long time suspended from caucus over a self-admittedly "not appropriate" relationship known generally in NT media as the "cocaine sex scandal" and then, very much later, kicked out of the broader party.   

Such polling as there is ...

Polling in the Territory has a reputation for being about as common as a vegan crocodile and doing not much better in the wild.  Not only are about six of the seats unpollable, but the small sample size in the remainder makes getting a representative sample seat by seat very difficult.  Despite this a uComms poll of Greater Darwin in 2020 proved to be very much on the money regarding the overall state of the parties.  

There may be more during the campaign but in the leadup we have so far had three polls.  A Redbridge poll in November 2023 had ALP 19.7% CLP 40.6% Greens 13.1% Independents 14% Shooters Fishers and Farmers 9.7% and Others 3.2%.  This poll was taken just at the start of the wave of COI scandals that eventually brought down Fyles; the 2PP might have been about 43-57.  The only poll since Lawler took over was a Freshwater Strategy poll released May 17 (date taken unknown) with Labor trailing 46-54 off primaries of ALP 29 CLP 39 Greens 9 and "independents" 22.  

Freshwater has had pretty good results at its few public tests so far.   Some hope for Labor against its numbers might be found in the federal poll's habit of being slightly stronger for the Coalition than other polls (but only by about 1%). Also the later of two private Freshwater polls for the Tasmanian election had a far worse result for Labor than occurred (that said this poll was still a few weeks out from that election).  

A less widely related poll (indeed I was not aware of it until reading comments sections on Tally Room just before finishing this article) was reported in the NT News on 24 June (link paywalled) and is a uComms of Darwin/Palmerston for the NT Environment Centre, but taken in March and April.  After redistributing initially undecided voters the poll had both majors on 35.7%, with a 57.5% flow of preferences suggesting that Labor would have a 2PP of about 52.1 for Darwin and Palmerston, a mere 2.1% swing against by my numbers.  While the 2020 uComms poll, also commissioned by an environment group, was very good, uComms polls can be very erratic.  The breakdown of other parties and independents in this poll is not available (I do know that Green, IND and others were the options offered, but not the support levels.) 

All these polls overestimate non-major voting options compared to what is likely to occur at the election, simply because of lack of suitable candidates.  The Shooters are not running at all (they are not even registered), the Greens have so far announced only seven candidates and tend not to run in every seat, and only ten seats so far (update Aug 4: twelve now) have independents.  So unless these numbers increase greatly, many prospective voters for these forces will probably not find anyone worth voting for and will go back to the majors.  In 2020 the Greens ran in 10 seats and recorded 4.4% of the Territory vote while independents recorded 10.7% from 17 seats.  

I may add notes on any further NT polls to this article or I may cover them separately.

Seat roundup

At the time of writing the Wikipedia article while highly useful has not been fully adjusted for the 2023 redistribution.  An estimate of notional margins based on the 2020 election can be found at The Tally Room.

Of the CLP's seven seats, Spillett (13.9%) and Nelson (22.8% vs ALP, 8.5% vs IND) are not going anywhere on a two-party basis.  Their remaining seats, Brennan (2.9%), Katherine (2.3%), Braitling (1.3%), Namatjira (0.3%) and Barkly (0.1%) are all very marginal on paper, but if there is a substantial swing to the CLP then all would be expected to hold easily, especially as the CLP has sophomore effect on its side in all of them ("double sophomore" in Brennan and Braitling.)  The CLP marginals are also not attracting much interest from indies, aside from Sam Phelan in Katherine.  

On the Labor side, there are a lot of complications.  Casuarina (16%), Gwoja (16.2%), Johnston (16.1%), Nightcliff (23.8%), Sanderson (18.8%) and Wanguri (17.3%) are on very high margins, though Wanguri is vacant and it will be interesting to see if Fyles' 2020-4 term dents her very large margin in Nightcliff.  (As a guide to how "safe" seats can fall in the NT when they're vacant, Barkly fell last election despite a margin of 15.8%).  

Some Labor seats are attracting some indie interest including former Darwin Lord Mayor Graeme Sawyer in Wanguri, David-Pocock-endorsed (!) Justine Davis in Johnston and Mililma May running against Fyles in Nightcliff.  I understand Davis to be running an especially prominent campaign with a high volunteer presence.  Yes the NT has teals too, but here they run in Labor seats and in cases with Gouldian finch logos!  Where I mention challenger indies in this piece I have no idea if they're competitive, I just point out some of possible interest. (I understand the Greens to be campaigning most vigorously in Braitling and Fannie Bay, both seats with no independents so far, as well as Nightcliff which is a traditionally strong seat for them).  

Further down the scale for Labor there are Fong Lim (2.1%), Drysdale (5.3%) and Karama (8.3%); these are all at varying levels of risk to the CLP if the swing is on and even more so is Port Darwin (2.1%) which is vacant and would hence have a good chance of falling even in a zero-swing election.  In Drysdale, Eva Lawler is likely to get some sort of vote boost for becoming Chief Minister, and might need it: Labor has only previously won Drysdale at the three elections where it won the 2PP.  

Now some even trickier cases, in alphabetical order:

Arafura (3.6%) is a remote Indigenous seat which saw a mid-term vacancy by-election.  Normally the low margin and the mid-term vacancy would make this seat likely to go on any swing but a caution about that is that incumbent Manuel Brown won the by-election with an enormous swing to him (15.6%).  A narrative here is that CLP mixed messages on the Voice to Parliament played a role, but I don't know if that's really true or not.  

Arnhem (1.6% vs IND, 17.6% vs CLP), another Indigenous seat, seems like it's the sort of seat at risk from an independent but not from the CLP, except the danger there is that the 2020 independent who nearly won, Ian Mongunu Gumbula, is now the CLP candidate.  

Daly (CLP 1.9%) is a seat disrupted by a mid-term change in ownership - in 2020 it had a CLP incumbent, it now has a Labor incumbent who won the by-election with a 7.3% margin.  Disrupted seats tend to behave somewhere between their previous margin and their by-election margin, but might behave more like the latter in the NT where personal vote effects are so large.  I'm considering it as more like a Labor marginal on, say, 5%, which is still risky.

Fannie Bay (10.9%) seems like it is on a hefty margin and should be safe unless the swing is very wild but was vacated mid-term and by Michael Gunner at that.  The seat has an erratic past history having been at times held easily by heavy hitters on both sides and could conceivably fall if Labor does badly.  

Now to the seats held by the crossbench:

Araluen (IND vs CLP 0.5%) - Robyn Lambley's margin in her Alice Springs based seat looks super precarious but nobody from the majors seems to be running against her so far, which seems ... really odd.  

Blain (ALP vs CLP 1.4%, IND-occupied) is a seat Labor had never won before Mark Turner picked it up in 2020, but with Turner now kicked out and the margin so small it would seem very challenging for Labor.  (For what it's worth they're not losing a personal vote since they didn't have one in 2020 either, but I suspect it's not worth anything much.)  One might think from afar that Turner's travails would seal his fate but a few weeks back the NT News referenced increasing views that Turner will survive, even citing a nameless CLP source calling him "a certainty'.

Goyder (IND vs CLP 6.7%, CLP vs ALP 14.4%) is former Speaker Kezia Purick's seat which she is retiring from.  Purick has endorsed Belinda Kolstad, a recently former CLP member and one of three independents running.  Purick's margin is not large enough to be sure if the endorsement will translate.  

Mulka (IND vs ALP 4.7%) - the CLP didn't run last time against Yolŋu independent Yingiya Mark Guyula who at this stage has attracted no challengers at all.  (Update: it appears that Labor will this time not run against Guyula while the CLP still say they are running.)

Overall 

Absent of any polls I would expect Labor to lose on average something like four seats here and probably just lose government compared to the 2022 election, just based off the NT behaving a lot like a state where federal drag is concerned.  Two of the three polls available are somewhat worse than that, while one is better.  A swing matching the one in the middle (the Freshwater poll) could potentially take Labor as low as six seats, though something like eight or nine would seem more likely for that swing based both on historic seat totals and on the seats where Labor would still have some arguably good chance.  To retain office, Labor most likely need to hold the swing to something more manageable, like, say, 4% (this is more than the swing in the uComms poll), and then to be lucky with the distribution in particular seats where there are arguments in their favour.  Uneven regional swing is also something to watch out for - Labor's best scenario would be a swing concentrated in Alice Springs where they don't have any seats left anyway.  What objective information there is mostly suggests that Labor deserve underdog status (currently a 40% chance in betting odds, which are not necessarily reliable), but not strongly enough to be all that confident of the outcome.  Hopefully there will be more polls before election day.  

Nominations Update

There are 80 candidates (25 CLP, 24 Labor, 11 Green, 20 independent/unendorsed.)  The 20 independents are running across 15 seats.  This is the fewest candidates since 2008 (65) which was also the last election with only three parties.  The number of indies might seem like a lot but the last two elections had more.  Labor is not contesting Mulka, which is handy for 2PP swing terms as it will be outside the 2PP for two elections in a row.  

See also

Tally Room guide 

Poll Bludger guide

ABC guide