Click here for full display of results.
Summary
Labor’s Joel Bowden finished the evening with a lead of 170 votes (2.6%) over Steven Klose of the Territory Alliance on the preference count, with probably only a few dozen votes outstanding. The Country Liberal Party humiliatingly finished in fourth place, with the Greens doing well to maintain their 17% share of the vote despite the expansion of the field from three candidates to seven. Together with the 21.4% slump in the Labor primary vote, the result is another illustration of the pronounced appetite voters have developed for minor parties when given a sufficient range of options. Another candidate, Braedon Earley, polled 10.4% on an anti-fracking platform, presumably benefiting from a considerable constituency hostile to both fracking and the Greens.
Labor was clearly buoyed by the strength of its candidate, a former Richmond AFL player and general secretary of Unions NT. While the Greens’ preference recommendation may have cut as much as 4% from the Labor margin, this was negated by the CLP’s equally curious decision to put Labor second, reflected in a near 50-50 split of their preferences (going off scrutineering figures obtained by Antony Green). A more normal CLP preference split would almost certainly have delivered the seat to the Territory Alliance — a result that perhaps scared the CLP more than a Labor win, potentially threatening their status as the main conservative party.
The result makes it very hard to determine how the August election might look. For all the strength of the Territory Alliance’s performance, its challenge in actually winning seats remains formidable — though perhaps not insurmountable if their near-success at the by-election inspires a bandwagon effect. The other notably strong performer, the Greens, do not enjoy the localised critical mass of support needed to win seats. However, it is clear both that voters are willing to turn away from the government, and that a CLP that can boast only two members of parliament is not seen as a credible alternative. The likeliest possibilities are either a bare Labor majority or a hung parliament with the Territory Alliance and independents as kingmakers or perhaps even coalition partners.
Live commentary
8.14pm. Eighty postal votes have been added — a bit more than Antony figured — on both the primary and two-party vote, breaking 47-33 to Labor.
8.11pm. My probability estimate assumes 304 outstanding votes, which is somewhat arbitrary. To the extent that that’s an overstatement, the remaining 1.9% probability of a Labor defeat disappears.
8.05pm. Rapid Creek EVC now in on two-party preferred, paring back my projection of Labor’s winning margin to 2.0%. They have a raw lead of 156 votes, with only 50 votes outstanding plus whatever the Darwin area mobile team will amount to, which probably isn’t much (and which I don’t expect will be favourable to conservatives).
7.57pm. Possible wild card: anti-fracking independent Braedon Earley’s preferences flowing heavily to the Greens, putting them ahead of the Territory Alliance, and after that who knows. But a long shot on both counts.
7.53pm. On Twitter, Antony relates there will only be about 50 postals, and a Northern Territory News reports Labor is about to claim victory.
7.47pm. Antony Green and I are of one mind: “I have three different methods for predicting the Johnston by-election result, and all three are predicting Labor to win with 52.6% after preferences.”
7.46pm. Rapid Creek EVC primary vote in, resulting in little change to the overall picture.
7.39pm. Moil booth has reported on two-party, behaving as my model expected to, leaving the Labor winning margin all but unchanged on 2.3%. Labor win probability now up to 98%, with the Rapid Creek EVC the only substantial unreported booth. However, this is a new booth that I’ve dealt with by dividing the results of the Casuarina EVC between the two, and it can’t be ruled out that its behaviour won’t quite be the same.
7.25pm. Millner two-party result in, with a slightly weaker preference flow from elsewhere bringing the projected Labor margin back from 3.0% to 2.4%. But because there are fewer votes outstanding now, this hasn’t changed my model’s estimation of a Labor win probability around 95%.
7.19pm. Moil booth added on the primary vote, and while it has the biggest primary vote swing against Labor so far, it hasn’t fundamentally changed the situation, projecting a 3.0% Labor winning margin. But for what it’s worth, the Labor win probability is back inside 95%.
7.00pm. The Casuarina pre-poll booth is in, on primary and two-party, and it’s firming up as a Labor-versus-Territory Alliance contest with the CLP still in fourth place. These numbers haven’t rocked my preference projection too hard, which is to say that Labor looks to be retaining a solid flow of Greens preferences despite the how-to-vote card. My model says Labor are very likely to win, with a projected 3.5% winning margin, but I’d still be conservative about interpreting it.
6.55pm. With that said, the Greens are doing well — their vote is up despite the fact they were the only minor party option in the field in 2016, compared with five this time, and they’re actually ahead of the CLP. So that slump in the Labor primary vote could partly be votes going to the Greens and coming back to them on preferences.
6.52pm. The Millner booth is now in — Territory Alliance continues to outpoll the CLP, while Labor’s primary vote has slumped by 20.7%. My projections remain rosy for Labor, but that assumes they will get 56% of preferences which I’m pretty sure won’t happen, because that’s calculated off a Darwin EVC result that had the CLP on 7.6%, whereas the Millner result is twice as much. So treat it with a grain of salt until we get the Millner two-party count.
6.49pm. Sorry, I had that the wrong way around — it’s Labor leading 40-26. And I think by results display is working now, and while it’s almost giving it to Labor, obviously you would want more numbers. Part of the equation here is that Labor got 16 minor party and independent preferences and the Territory Alliance got 11, which needless to say isn’t much to go on.
6.40pm. The two-party count for Darwin ECV bodes well for the Territory Alliance, who lead 40-26 — but not for my results display, which has tanked under the pressure of having the parties other than I expected them to be. Will see if I can fix.
6.28pm. The NTEC’s results are on display now here, and we have 66 votes from the Darwin pre-poll centre. Obviously that’s not much to go on, but it’s interesting that the Territory Alliance has 15 votes to the CLP’s five. So far so good for my own results display — I’m projecting a 36.6% primary vote for Labor, which suggests they’re in trouble in less preferences behave in an unanticipated fashion.
6.20pm. Hopefully the plan is for the NTEC results display to come to life when there is actually a result to report. I will try to just swap Territory Alliance for CLP in my two-party calculations, so the latter’s historic results are used to calculate the swing for the former, but I don’t know how smoothly that’s going to run. That’s assuming that the NTEC is planning on publishing anything …
6.16pm. Still no sign of any results facility on the NTEC site, and no media feed in operation. I asked Antony Green on Twitter where he would be getting his results from, and his answer was “I don’t know yet”.
6pm. And they’re off. Antony Green relates on Twitter that the Northern Territory Electoral Commission have surprisingly decided to make their indicative count between Labor and the Territory Alliance, which means I won’t be able to calculate two-party swings. There’s also no sign of any results display on their website.
Preview
Today is the day of the Northern Territory’s Johnston by-election, which also happens to the first election of any kind in Australia above local government level since the federal election last May. Labor holds the northern Darwin seat by a margin of 14.7%, but the seat is less secure for them than this makes it appear owing to the scale of the Labor landslide in 2016 and the importance of local and candidate factors in the territory’s boutique electorates, which have only around 5000 voters.
A very substantial swing against Labor can be anticipated due to the departure of sitting member of Ken Vowles and his estrangement from the party, together with the general difficulties that have beset Michael Gunner’s Labor government since it came to power in 2016. There is also the fact that the Greens are instructing voters to put Labor last in protest over the government’s lifting of a moratorium on gas fracking exploration, although the effect of this is limited by a prohibition on disseminating how-to-vote cards near polling places.
All of this bodes well for the opposition Country Liberal Party, although they face opposition for the conservative vote from Territory Alliance, a new party founded by former CLP Chief Minister and now independent MP Terry Mills which is making its electoral debut. For what’s it’s worth, the latter’s candidate is the $1.70 favourite at Sportsbet, which is offering $1.90 for Labor and $2.75 for the CLP.
Live coverage will follow here upon the closure of polling at 6pm, encompassing analysis on this post and a detailed display of results that is ready to go here. Naturally though, in an electorate this small there are only so many results to follow – two election day polling booths plus two pre-poll booths, with the latter accounting for an ever increasing share of the action.