NSW 2015 – introduction to the Legislative Council count


I’ve waited until now to post about the Legislative Council, since the count is likely to take some time, and so many votes have yet to be counted. But at this point I think we can start to analyse what is going on in this count.

As of last night, just under 2.9 million votes have been counted. There are slightly more votes included in the count on the ABC website than on the NSWEC website (just over 2000 votes, at the time of writing) – I’m not sure why that’s the case.

It’s worth noting that the NSWEC at the moment is including blank ballots and ‘other’ ballots in their calculation of vote percentages. Blank ballots are obviously informal (although they will be checked again), so they can easily be discarded when calculating the position of each party. ‘Other’ ballots include some ballots that may be informal or where the intention of the voter isn’t immediately clear to the person counting on election night, but they also include all below-the-line votes.

On the ABC website, percentages are calculated by excluding other and blank ballots, so the same raw vote figures will produce higher percentages for all parties.

Once the first round of counting is complete for all above-the-line votes, and all ballots are sent to the central counting centre for data entry, we will start to see fully-processed votes including below-the-lines being reported. At that point it will be possible to judge:

  • What proportion of those ‘other’ votes are informal, as opposed to below-the-line votes
  • How below-the-line votes skew compared to the above-the-line votes already counted in places where below-the-line votes are yet to be counted. In 2011, we found that below-the-line votes significantly favoured the Greens and were poor for the Liberal Party.

The other problem is that there are a lot of booths where no Legislative Council votes have been counted, as well as most categories of special votes (such as absentee, postal, pre-poll votes).

In addition, the progress of the count in each seat is at different points. In the seat of Balmain, only 4 out of 33 booths (counting categories of special votes as a booth for this purpose) are classed as ‘Initial Count Complete’, while that figure is 43 out of 53 booths are in that category in Bathurst.

As a quick measure, there is a slight positive correlation between percentage of booths that have reported upper house vote and the Liberal vote in the lower house, and a negative correlation with Greens vote. This suggests that Liberal booths are over-represented, and more pro-Green booths are under-represented.

With those massive caveats, this is where the current numbers stand. All of these numbers basically exclude any possible informal or below-the-line votes.

The Liberal/National coalition has won nine seats, Labor has won six and the Greens have won two, all on full quotas, and there seems to be no chance any of these seats will change hands.

In addition, Labor has 95% of a seventh quota. While that could go up or down, Labor should comfortably win a seventh seat. The Shooters and Fishers likewise have 85% of a quota and will comfortably win a seat.

Fred Nile of the Christian Democratic Party is on 61% of a quota. This compares to 68.7% of a quota in 2011. Nile is probably safe on these numbers – no candidate on more than 53% of a quota has failed to win.

So after electing nine Coalition, seven Labor, two Greens, one Shooter and one CDP, that leaves one remaining seat to be decided.

At the moment, the tenth Coalition candidate is in pole position, sitting on 47% of a quota.

Other candidates with a chance are No Land Tax on 0.37, Animal Justice on 0.35 and the third Green on 0.20.

It seems likely that the Coalition vote will drop and the Greens vote will rise as the remaining election-day votes are counted, and this will go further as below-the-line votes are included. It’s much harder to know whether No Land Tax or Animal Justice will go up or down as the count becomes more complete. The Greens vote will need to rise substantially for them to be a serious threat, but the outstanding factors should all favour the Greens.

In the next few days we should have a much better sense of who is in contention, and we can begin to predict how the remaining below-the-line votes and special votes will break.

One other thing worth noting: there were 19,000 iVotes cast early in the voting process which were missing an above-the-line box for the Animal Justice Party and the Outdoor Recreation Party. While those voters had the opportunity to change their vote, there could well be grounds for a legal challenge if the AJP come close to winning a seat, and miss out.

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


  1. If Animal Justice pick up Greens preferences, pass No Land Tax and end up behind the Libs by less than 200 votes we might be back to the polls.

    Because if Animal Justice is behind the Libs, but it can be shown that if those 19000 iVotes were representative of the other iVotes then Animal Justice would win, the court is in a sticky situation.

    Lets hope it doesn’t come to that, but based on the numbers you’re throwing up it might just.

  2. I don’t think you can do that, it creates a very uneven playing field.

    It was also possible to track down all the voters who voted at polling places in WA that lost Senate ballots, but it was unacceptable to just let those booths vote again.

  3. Absentees tend to favour the Greens and they are yet to show, but pre-poll and postals go against – are they loaded? The Greens need to creep up to 0.4, then hope for Socialist Alliance, Voluntary Euthanasia, Cycling Party, Democrats preferences to pass Animal Justice, and hope for 0.1 of a quota from AJP preferences. It’s a hard to slog, but not impossible. I don’t think Nile is home and hosed either

  4. You have to be able to vote with the same knowledge as everybody else. Knowing that candidate A only needs x votes to beat candidate B is not the same knowledge the rest of us voted with. You could change your vote from (or to) a 1 for candidate C just to affect the outcome (actually its only really to do this in optional preferential wheras in compulsory preferential your preferences would already flow).

  5. I was old before I ever voted. This my last election has been expensive because i publshed four newspaper pages intended to influebce the outcome of the vote in Gosford and is really tedious because its taking ages to get a result! Why? Edward James on the long paddock

  6. “Fred Nile of the Christian Democratic Party is on 61% of a quota. This compares to 68.7% of a quota in 2011. Nile is probably safe on these numbers – no candidate on less than 53% of a quota has failed to win.”

    What you intended is: “no candidate on MORE than 53% of a quota has failed to win.”

  7. on the Animal Justice Party not having an above the line box for the first 19000 votes cast – they are tracking to get about 1.7% of the vote. Presumably they would argue they were denied 1.7% of those 19000 votes or about 330 votes. If the final margin they miss out to the 10th Liberal is less than that they have appear to have a good case – if the margin is much greater, say over 1000, the case I would think has much less merit.
    On current projections they are on 0.37 quotas and will get some of the Greens surplus (currently 0.11 quotas) for voters who followed a how-to-vote. My guess is about half at best – so final would be 0.42 quotas well behind Libs on 0.54 quotas. I discount No Land Tax as they don’t benefit from any decent preference flows.
    If this is about right the margin of 0.10 quotas would be about 19,600 votes (estimating final quota at 196,000).
    It’s a long way to go but I doubt legal action will eventuate

  8. The ABC’s count just reached 4.0m, and since election night the 10th Liberal has risen from 47% of a quota to 53%, while the 3rd Green has dropped from 20% to 10%. Ben, is that a surprise given that you thought the election night figures included more Liberal booths and fewer Green booths?

    Obviously below the line votes haven’t been counted at all, and the figures could shift noticeably when those are added in.

  9. Most booths yet to report absentee votes …
    (only can see for Gosford, Strathfield and The Entrance)

    While there are just small numbers of votes [~500 total formals] showing on the NSWEC vtr (perhaps just thus far?), here the Greens have respective increases in %of formal votes of 8.7%, 6.58% and 5.17% [average: 6.82%] compared to the total electorate counts, leading to absent % of 17.43%, 15.34%, 14.17% respectively. These are all electorates where greens vote is at ~9%.

    Wonder if that will be maintained/replicated across the state? [lismore]

    Wish we knew when all the other absentees would be counted/entered to help estimate for some of the marginals!?

  10. Also (on those ~500 absents in each of those electorates):

    Showing figures of {absent votes/% of formal votes/% difference b-w PARTY total formal votes and PARTY absent votes}

    20 3.38% 1.55%
    249 42.13% -0.80%
    14 2.37% 0.18%
    171 28.93% -9.92%
    103 17.43% 8.70%
    34 5.75% 0.30%

    87 15.34% 6.58%
    17 3.00% 0.77%
    224 39.51% -2.82%
    29 5.11% 1.90%
    210 37.04% -6.44%

    162 45.00% 0.68%
    8 2.22% -0.44%
    51 14.17% 5.17%
    123 34.17% -7.82%
    16 4.44% 2.41%

  11. *oops – should read:

    Showing figures of {absent votes/% of ABSENT formal votes/% difference b-w PARTY total formal votes AS % OF FORMAL VOTES and PARTY absent votes AS % OF ALL ABSENT FORMAL VOTES}

    [hopefully you get the drift]

  12. A comment on how fast or slow the election count is. The only date that matters to the NSWEC is the date on which the writs must be returned, which is 11 May, a good five weeks away. They are not conducting the count to satisfy the curiosity of election buffs, or even candidates. Their duty is conduct the count carefully and accurately, not quickly. We saw with the WA Senate election how important that is. The count should take as long as it needs to take.

  13. I agree.

    I’m working on another LC post for tomorrow morning which lays out how many above-the-line votes are yet to be counted (mostly absentee and postals) and the possible shape of the below-the-line vote.

  14. I notice that the dogs of MSM (or RN Breakfast at least) have started barking about NLT and the role of the LNP in their apparent success plus the possibility of NSW returning to the polls for the LC. Your thoughts on this would be illuminating.

Comments are closed.