NSW 2015 – projecting the Legislative Council


There’s a lot of outstanding questions in the Legislative Council count. We know that the current vote shares will shift, and as they shift, these could have a significant impact on the chances of any party to win a seat.

Using estimates of how many votes are outstanding, and looking at how parties performed in those vote types at the last election, it’s possible to make a projection of how the vote percentages for each party will shift, with the addition of absentee, postal and below-the-line votes over the next few weeks.

There are two main effects that will change the votes for each of the candidates:

  • Proportion of the vote received in the ‘special votes’ – mostly postal and absentee votes.
  • Proportion of the ‘below-the-line’ vote received for each party.

It is reasonably clear who will win the first twenty seats: nine Coalition, seven Labor, two Greens, one Shooter, and Fred Nile of the CDP.

When factoring in current trends, I can make the following projection for where each party will stand at the end of the count.

PartyCurrent votesCurrent %Current quotasProjected %Projected quotas
Christian Dem.103,6532.75%0.60552.74%0.6022
No Land Tax66,0541.76%0.38581.76%0.3870
Animal Justice63,0601.68%0.36841.66%0.3657

On current figures, it seems like #10 Coalition candidate Hollie Hughes (a Liberal) is well in front of her main rivals, Peter Jones of No Land Tax and Mark Pearson of Animal Justice.

Yet when you look at the prediction, the Liberal vote drops from 54% of a quota to 41% of a quota, only 0.0222 quota ahead of Jones and 0.0435 quota ahead of Animal Justice.

While the Greens vote picks up substantially, #3 Greens candidate Justin Field would be well behind the leading candidates in this scenario, which already assumes a significant increase in the Greens vote on late counting.

If this projection proved true, the best-case scenario for the left would be that Greens preferences would flow to Animal Justice (as indicated on their how-to-vote) and elected Mark Pearson ahead of Hollie Hughes or Peter Jones.

Of course, if Animal Justice falls just short of winning, it’s possible there could be a legal challenge to the result, because of the 19,000 iVotes which didn’t have an ‘above-the-line’ box for Animal Justice or Outdoor Recreation.

As the count proceeds, most of the above-the-line votes will be counted and the NSWEC will begin conducting their data entry of below-the-line votes, and we will be able to make more precise projections over the coming weeks.

Hopefully it will also be possible, through scrutineers, to get a sense of how preferences are flowing, to determine whether the primary vote figures could be overturned by preferences, in particular whether Animal Justice can win on Greens preferences.

Below the fold, I will run through a bunch of stats I’ve been able to pull together, and explain how I have used these figures to produce my projection.

This Legislative Council count effectively goes through two stages. Firstly, all votes are sorted – above-the-line votes sorted by group, and below-the-line votes and informal votes grouped together as ‘others’, and blank votes also grouped together. In this count, below-the-line votes are simply bundled together with informals, and are not split up by the candidate who receives the vote. This stage is well under-way.

At some point, these ballot papers are gathered at the central counting centre in Riverwood, and data-entered. As this process proceeds, a second count of votes is released, which will include below-the-line data. At this point we will be able to run comparisons between the first count and the second count to identify which votes are missing, and project how the remaining below-the-line votes will flow based on those counted so far.

The following table shows the final number of votes cast in 2011 by vote type, and the number of votes counted so far by vote type (including blanks, informals and below-the-lines).

Vote type2011 final2015 so far
New enrolment21,3770
Declared institution13,371

In 2011, iVotes were included with postal votes, and declared institution votes were included with pre-poll votes.

As it currently stands, Legislative Council above-the-line votes have been reported for all ordinary votes, except for two booths in Miranda and Willoughby districts.

There are also only eight pre-poll booths that haven’t been reported. Overall, there has been an increase of about 60% since 2011, which may explain why the number of ordinary votes have dropped.

The number of declaration votes counted in 2011 during this first count was 1.1 million, but this time only 425,000 votes have been counted. Two thirds of these votes are iVote electronic votes, along with over 100,000 postal votes. No absentee votes have been counted yet, whereas in 2011 over 400,000 absentee votes were counted, and there should be a lot more postal votes to come.

So we should still be expecting a bunch more above-the-line votes to be counted over coming days, mostly absentee and postal votes.

Then we have the other major unknown factor in the race at this point: below-the-line votes. So far, 153,000 ‘others’ votes have been tallied up. This compares to a total of 188,032 others votes tallied up in 2011. Out of these 188,000 votes, 91,428 votes turned out to be formal below-the-line votes in the final count.

One big outstanding factor in 2011 was the presence of Pauline Hanson. Since she ran as an independent, she didn’t get a name above the line, and received a remarkably high number of below-the-line votes. There is substantial anecdotal evidence that many Hanson voters just voted ‘1’ for her below the line, and thus their votes were informal.

Having said that, our best evidence suggests that about half of those ‘other’ votes being tallied up are formal below-the-line votes. If you assume some more are yet to come in, it would be reasonable to think the total number of formal below-the-line votes is somewhere between 80,000 and 100,000.

The following table outlines the 2011 vote for each of the parties in Parliament in each of the vote types that haven’t yet been counted, along with an estimate of the number of votes yet to be counted in each category.

Votes remaining450,000180,00020,00020,000100,000
Christian Dem.3%3%3%3%3%5%

I have adjusted below-the-line votes to exclude Pauline Hanson votes, since I think those were an anomaly caused by her not having her name above the line.

You can see from looking at these figures that the Greens should do well from a majority of the votes that are yet to report, while the Coalition should do badly. This explains my previous assumption that the Greens vote could rise to a competitive level on late counting.

So using these figures, I can then produce a competitive model, allocating these outstanding votes to each party in proportion to their vote, taking into account their 2015 ordinary vote, and the difference between their 2011 ordinary vote and their 2011 vote in that vote type.

When I finish that process, I get the prediction I produced at the top. In this prediction, the Liberal vote drops, while the Greens vote increases. The change is probably not enough to put the Greens in with a chance, but it is enough to make the Liberals nervous about missing out, in favour of No Land Tax or Animal Justice.

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


  1. Ben, is it correct that the NLTP didn’t manage a preference deal with the Libs and so are likely to drop below the AJP in the final count?

  2. I think that’s right. The NLT how-to-vote (at least in the inner-city) was aggressively anti-major parties and instructed people to just vote 1 (and emphasised not preferencing). Combined with the reports in the media about conflict between the Liberal Party and Peter Jones, there seems to be decisive evidence that they had no deal.

  3. Tony Nutt was threatening to disendorse LA Libs candidates who did preference swaps with the NLT counterparts, so I think it’s safe to say there was no state wide direction from the coalition to preference NLT.

  4. If any Greens preferences do flow to AJP which is probable. The final count is likely to go between the libs and AJP, depending on how many preferences the libs get from other parties, who do you think will get the final seat?

  5. In Gosford the two parties not much preferred both indulged in illegal campaign activities holstien had two core flutes nailed back to back photographed back on power pole in Woy Woy. They are reported to have been displayed influencing votes for five days just like the three core flutes on the road reserve of Brisbane Water Drive. I ont know about other constituents but I expect candadiates to conduct their campaign within the law! Edward James on the long paddock.

  6. Edward, Labor had huge banners up on public land next to Blackwall Rd Woy Woy featuring the typical lame Labor scare campaign. I wish I tore it down.

  7. From talking to a Greens colleague, she says that Greens scrutineers are seeing tight preference flows to the AJP as well as preferencing below the line- so it will be a nailbiter to the end.

  8. NLT and AJP are running roughly equal across the state. However here’s something iv’e noticed-

    The ethnic diversity of the seat has a lot to do with whether NLT or AJP perform better.

    NLT performs much better than AJP in the very multicultural seats like Liverpool, Auburn, Strathfield & Rockdale

    AJP meanwhile outperforms NLT in the predominately Anglo-Saxon seats (regardless of them being Lib/Lab or Syd/Country). Good examples are Hawkesbury, Tamworth, Pittwater, Camden & Monaro.

    Its the ethnic diversity of the seat which is largely deciding whether NLT or AJP perform better. (I think its mostly to do with ethnic voters being far more likely to donkey vote in their LC vote).

  9. Latest figures has the following seats locked in:

    Coalition: 9
    ALP: 6
    Greens: 2

    Total: 17

    Assuming that any party with left-over quota above 0.75 is pretty much guaranteed a seat, that then gives the following:

    ALP: 1 Seat
    Shooters and Fisher: 1 Seat

    Total: 19 seats

    This leaves 2 seats up for grabs. Taking into account anyone with 0.25 quotas or better:

    CDP: 0.61
    Coalition: 0.58
    NLT: 0.38
    AJP: 0.37

    From the last election, the CDP was elected from 0.68 quotas and the 11th person on the Coalition ticket was elected with 0.4885 quotas, with the Greens gaining a seat from 0.4457 quotas. The highest left-over quota not to get a seat was Pauline Hanson’s group who got 0.52 quotas.

    This time, the Greens will have 0.08 left over for their quotas so they are not in the running. Of these 4 parties, you have 3 from the right (assuming that No Land Tax is a right-wing party) and 1 left wing Party (AJP).

    The AJP will need a strong flow from the micro-parties (the kind of which saw Ricky Muir elected in Victoria).

    Based off last year, my money would be on the CDP holding on and the Coalition scraping the last seat just ahead of NLT and AJP (with preference exhaustion in the Upper House playing a major role). But if I was to put money on anything else happening, it would be on the AJP getting elected on a strong preference from the micro-parties and sneaking ahead of the Coalition (Assuming they don’t manage to jump Fred Nile).

  10. Agree Greens are not in the running but their over two quotas amount is creeping back up and will probably rise with below the line votes – any improvement there will probably improve the AJP chances

  11. Further modelling with regards to the recent dispute regarding the potential challenge by the Animal Justice Party, who have threatened to take the result of the LC to the Court of Disputed Returns if they don’t win a seat over the failure of VoteNSW to include the AJP and the Outdoor Recreation Party on 16,000 iVote Ballots

    Currently, AJP are on 0.37 quotas whilst the Coaltiion have dropped to 0.55 leftover. With the AJP on 67,727 votes (giving 0.37 quotas) and the Outdoor Rec Party on 28,167 (0.15 quotas), the Coalition’s 0.55 quotas goes to the equivalent (current) of 100,000 votes, meaning that there is currently a 32,000 vote gap between the AJP and the Coalition for the last seat.

    Assuming a uniform spread of votes, that would mean that the AJP would have received 270 votes and the Outdoor Rec Party would have received 112 votes. The AJP would then be up to 68,000 votes, giving approx a 31,000 vote gap.

    It really comes down to in what order the parties are eliminated for their quotas. The first critical point will be the 10th elimination of parties. It then becomes a race for the last two seats. I think whoever finishes 2nd between the CDP and the Coalition will be directly in the firing line of the preference deals going on under-neath.

Comments are closed.