Upper Hunter by-election live


10:14 – It’s looking a decent-sized swing to the Nationals on the two-candidate-preferred vote. The existing margin was 2.6%, but it’s looking like the new margin will be something like 7%, bearing in mind that a lot of pre-poll votes are yet to report.

I’ve broken down the booth results kindly compiled by William Bowe, into the same geographic categories I used in my pre-election guide (mostly following local government boundaries).

The Nationals two-candidate-preferred swings by area are:

  • Dungog – 13.2%
  • Gloucester-Stroud – 10.3%
  • Liverpool Plains – 3.8%
  • Muswellbrook – 10.1% to Labor
  • Singleton – 8.2%
  • Upper Hunter – 1.1% to Labor

And with that I’m finished for the night. Thanks everyone for joining in. I’m now going to go dark for a little while as I prepare for the federal election, before returning with extensive coverage of the NSW local government elections.

10:01 – And here is a map showing the primary vote for the five top-polling candidates.

9:36 – Sorry I’ve been quiet for a while but now I’ve got a map showing the two-candidate-preferred results and swings by booth for the election-day vote.

Muswellbrook acted very differently to the rest of the seat, with Labor gaining large swings across that council area. Labor also had slight gains in Scone, while the Nationals went up across the east of the council area.

8:37 – We are waiting for a handful of election-day booths to report their preference counts, but overall it looks like big swings to the Nationals in the Dungog and Gloucester-Stroud areas, big swings to Labor in the Muswellbrook area, and relatively smaller swings in the other areas, but they also favour the Nationals in the Liverpool Plains and Singleton areas.

7:49 – I will do a full map later tonight and I’ll include a breakdown of the vote by the same sub-areas I included in my pre-election guide, but as a quick preview, the One Nation vote is highest in Singleton (18%), while the Shooters vote is much more evenly distributed. At the moment the swings to the Nationals by region all look quite large at the moment but I think they’ll shrink once the big town booths report.

7:41 – The overall two-candidate-preferred swing has been alternating between Labor and the Nationals but either way it is slight. The current Poll Bludger projection has the Nats up by 0.5%.

7:28 – The primary vote gap between the major parties is growing, with the Nationals up to 34.7% and Labor down to 18.9%. That’s just seven percentage points ahead of the Shooters on 11.9%, but that’s probably still too much for the Shooters to close.

7:17 – By the way, the best place to see the results is at the Poll Bludger results tool.

7:16 – We’ve now got two-candidate-preferred counts from seven booths, overall the Nationals are up 1.3%. About two thirds of all preferences from candidates other than Nationals or Labor are exhausting. On those numbers it looks like the Nationals will win, but I wouldn’t call anything yet.

6:57 – It’s hard to see any of the independents or minor parties making the top two. The Shooters would need to basically gain every One Nation preference, plus a few more, to overtake Labor (assuming Labor doesn’t gain any preferences). Remember that optional preferential voting means some preferences exhaust, and means you need to do better on the remaining preferences to catch up to a leading candidate.

6:56 – Worth noting that the booths in so far are all quite small. We have one out of three booths from Singleton, and one out of three from Scone, but votes haven’t come in from Muswellbrook or Dungog.

6:54 – A handful of booths have reported and at the moment the Nationals are leading with 32.7% of the primary vote, followed by Labor on 22.2%, the Shooters on 11.5%, independent Kirsty O’Connell on 10.9%, and One Nation on 9.8%. It’s far too early to call anything.

6:00 – Polls have just closed in the NSW state by-election for the seat of Upper Hunter. This is a key contest in a very marginal seat, with the Nationals under threat from Labor, One Nation and the Shooters, Fishers and Farmers party. You can bide your time before the results come out by reading my guide to the by-election, which features at least 44 comments as of Saturday afternoon. I’ll be back around 7pm with the first results.

Liked it? Take a second to support the Tally Room on Patreon!
Become a patron at Patreon!


  1. Doubt we will have a clear winner tonight. There is no way the Nats win it with well under 35% primary predicted. This could be very messy although it is early in the count

  2. Daniel, what do you mean “no way the Nats win it”?? Nats are on 33%, and over 10% ahead of Labor on primaries. Everyone else is barely pushing double figures at best. I’d say they are highly likely to win bar some miracle preference cascade.

    It’s OPV so much harder for someone to overtake Labor into second. According to Antony Green/NSWEC, around 65% are exhausting.

  3. Well I’m certainly not conceding yet as much respect I have for Antony. Name instances where candidates won a seat with 30% or less of the primary vote. It rarely happens. Where are the preferences coming from? One Nation? Why would One Nation who are Pro-gun preference the Nats over the Shooters?

    And Barilaro’s cheepshot at Malcolm Turnbull was totally unacceptable if there is anyone who should be expelled from one of the coalition parties it should be John for political treachery for preferencing Mike Kelly in 2019 over Fiona Kotvojs. Malcolm maybe retired but isn’t gone.

    This electorate might be more friendly to coal but this by-election was more about the pandemic like almost all elections are during this period, In 2023 this could be a Shooters gain if they recover or if Labor fix themselves they could gain this but the Nationals shouldn’t be complacent about holding this forever even though they are likely to hold it this time (Although I am not conceding JUST yet)

  4. I suggest you check out the results facility on Poll Bludger. They’ve already distributed preferences between Labor and Nationals and the Nationals are clearly winning. They don’t need many preferences because 2/3 of other candidates’ voters are exhausting. And it’s not relevant how One Nation voters preference between the Shooters and Nationals, since the Shooters came third.

  5. I wasn’t aware NSW still used OPV, I thought we were the final state (QLD) to abolish it.

    It now looks clear the Nationals have won basing it on the amount of exhausted votes, The government should consider a referendum on CPV to prevent such a high exhausted vote happening again. There are positives and negatives for CPV and that would be for the people of NSW to decide, It has been used federally I believe since 1984?

  6. The Labor government introduced OPV in 1980 so they it would hurt the Coalition in 3 way contests, but it has backfired on them. It is unlikely the Coalition will repeal it as over the past decades due to how much it has benefitted them, the Coalition’s views have changed and they now favour OPV.

  7. Labor still wouldn’t’ve won this under CPV – assuming 80% of Green preferences and half of the others (that’s being generous) would get them about 46% 2pp. CPV would’ve increased the chances of the Shooters, One Nation or even O’Connell jumping over Labor and into a 2cp count, but the Nats would still be favourites there.

  8. Yes opv seems to benefit the libs and nats. The gap was 10% on primary votes.. This came back to about 5% after preferences add the exhausted votes back in and it would have been very close. Maybe shooters may have been in the mix I don’t know
    There was a poor national vote as well only 30%

  9. But that’s the system Mick and Daniel. Bought in by the ALP in 1980 or so. And it was the system when both of you triumphantly stated that the Nats couldn’t win.
    As I’ve said before Mick, predictions are always better if based on what you think might happen, on an educated basis, rather than what you hope might happen.

  10. opv tends to benefit the party travelling the best…. makes it hard to catch up from behind.
    moderate you are just as biased as me in the opposite direction….. have you ever picked labor to win.
    the nats were very lucky to win……. the logic of a byelection caused by misbehaviour suggests a loss…. the govt is reasonably accident prone……. np short term move to the back bench, Gladys bad memory about her corrupt partner. one minister being investigated by icac, another by the police.

  11. Mick, exhausted votes can’t be expected to be anything apart from a 50/50 split between the two majors (if you forced them to number all boxes). No preference either way means exactly that. Artificially adding them back in shouldn’t change a 2pp result.

    The sort of seats Labor win under CPV but lose under OPV are where there’s a strong Green vote and few minor right parties. Example: Moreton last federal election. Labor started out 5.7% behind the LNP, and of the 24% for minor parties, two-thirds of it was for the Greens – Labor won by 1.9%. Seats like that are why the Libs want OPV federally.

  12. It would probably help if Labor at least tried to offer some convincing argument why they’d be substantially better than the Coalition on issues that Greens and other minor left-wing parties care about, but obviously Albo’s entire strategy has been to be as bland and nonconfrontational as possible and somehow sneak into power on the back of a timely scandal or whatever. If the Libs introduce OPV before the federal election you can practically call it for them now.

  13. One very faint silver lining is that the Greens will almost certainly be talking up their chances of winning balance of power in a Labor minority government, and therefore will probably direct preferences to Labor on HTVs. Still won’t even come close to evening the odds of course.

  14. What’s the exhaust rate for O’Connell votes? Her anti coal position would have gotten primary votes from the left (thinking not pro-coal ALP voters who don’t like the Greens). Her HTVs didn’t recommend preferences and I imagine voters would interpret that as “just vote 1” (especially with all the official looking “Just vote 1” signs at booths).

    Anyway incumbency will probably keep this in Net hands next election even if Labor are competitive. The pro coal signalling might have actually worked (Muswellbrook) but Labor is unlikely to repeat that in a statewide campaign.

  15. The full distribution of preferences and exhaustion rate per candidate has not been published yet, as there are still some postal votes and provisional ballots to be counted. Currently they are being entered electronically, however they can only be completely entered when all votes are in. Then the exhaust rates and the distribution of preferences will be calculated electronically. All of this will be done by the end of the week.

  16. The exhaust rates aren’t out yet, they will be entered electronically after all provisional ballots and remaining postal votes are counted. The exhaustion rates will then be calculated electronically and will be published by the end of the week.

  17. Bird of paradox… I see your point but if a 50/50 split was the case how would not this be exactly the same as opv. look at votes of non majors. 8% odonnell 5% drayton 12% sff 12% onp and 3% greens…… 42%…. this would benefit the alp better than 50/50 and if push comes to shove… maybe sff…. if they leapfrogged labor. I accept your point that I over simplified in trying to measure

  18. Here’s an example: Under CPV, Greens preferences can be expected to go 80 ALP / 20 Lib, but under OPV it’d be more like 40 ALP / 10 Lib / 50 exhaust. The proportion of non-exhausting prefs between the two major parties is the same as under CPV (*), but the number of votes involved is less, which makes it harder for whoever’s first on primaries to lose.

    (*) At least, I assume so? I could be wrong on this, but comparing NSW with either other states or with federal results is tricky.

    I doubt the minor party vote in Upper Hunter would favour Labor. The Greens and Sustainable Aus would, but they only got 4% between them. One Nation got 12% and are strongly anti-Labor, and I can’t imagine the rest would be anything more than 50-50.

  19. Based on an estimate of FPV in NSW 2019 the following seats would have had a different outcome
    Dubbo: IND gain instead of NAT hold
    East Hills: ALP gain instead of LIB hold
    Penrith: ALP gain instead of LIB hold
    Wollondilly: IND gain instead of LIB hold

  20. Balmain would have been an ALP gain from GRN in 2015 under full preferential, and 2019 would have been close. In FPV, Liberal voters preference Labor over Greens (although they went the other way in 2010 when HTVs told them to). In OPV, they just vote 1.

Comments are closed.