NSW 2019 – what about the upper house count?

The Legislative Council count is going to take a long time, and we are missing a lot of vital information we will need.

The main reason for this is the Electoral Commission’s decision to only count the votes of seven big groups on election night and in the first round of counting. But there are other reasons which apply in every Legislative Council count. Below the line votes are mostly not counted in the early parts of the count, and we don’t have any hard evidence about how preferences will flow (unlike in the lower house, where two-candidate-preferred data answers this question in most seats). And finally, a much smaller proportion of votes have been counted in the upper house compared to the lower house.

I can’t answer all those questions, but in this post I will try and shed a bit of light and give a sense of when we can expect answers.

First, let’s look at the statewide voting figures declared so far. First, let’s restate that the ‘others’ vote includes all below-the-line votes, all informal votes with markings, and all above-the-line votes for any group other than the seven groups who have been counted. There is a separate category called ‘blanks’ which we can safely ignore as informal.

Before I start, it’s worth noting that this the figures below are based on just over 3.3 million votes having been counted. 5.27 million people are enrolled to vote in New South Wales, and almost 3.9 million votes have been counted so far in the lower house, so there’s a lot of votes yet to be counted, even in the rudimentary way votes were counted on election night. I won’t try and dive into the question of where these outstanding votes are, but that could be a topic for a future post.

I’m going to use the same calculations as Kevin Bonham, who took the raw figures, then discounted 1.5% of the marked ballots as informal and then slightly increased the vote of the seven main parties in line with their traditional ratio of below-the-line votes to above-the-line votes. I will simply use these same formulas but with the latest voting data. All numbers are based on quotas – a quota is 4.55% of the total formal vote.

All of the assumptions applied here effectively reduce the size of the ‘others’ pile, and thus increase the quotas for the seven parties named.

At this point it is safe to assume that the following 17 seats have been decided:

• 7 – LNP
• 6 – ALP
• 2 – GRN
• 1 – ON
• 1 – SFF

And then the next candidates in order with a chance at winning the last four seats are:

• 8th LNP (0.5958)
• CDP (0.4700)
• 2nd ON (0.4064)
• AJP (0.4028)
• 7th ALP (0.3236)

But of course this isn’t the full picture, since there is about 2.5 quotas sitting between 13 other groups:

• Socialist Alliance
• Sustainable Australia
• Advance
• Chris Osborne
• Flux
• Jeremy Buckingham
• Conservatives
• Keep Sydney Open
• Liberal Democrats
• Voluntary Euthanasia
• Christian Democratic Party
• Small Business Party
• James Jansson

It seems likely that Socialist Alliance, Advance, Chris Osborne, Flux and James Jansson will all get a very small vote, leaving 2.5 quotas to be divided between nine other groups. That would average out to 0.28 quotas per group – not far off the 0.4 quotas which has the Animal Justice candidate currently on track to win a seat.

So if you assume that this vote is not evenly distributed, it’s easy to imagine a number of these groups could be up around 0.4 or 0.5 quotas, and in the race for the final seats.

So what do we have to go on in judging which of these parties have the best chance? It’s worth noting that Sustainable Australia, Animal Justice and Keep Sydney Open all ran candidates in a large number of seats (55, 48 and 42 respectively), and are all sitting on 1.5% of the statewide lower house vote at the moment. This lower house vote has translated into 0.4 quotas for Animal Justice, which suggests Sustainable Australia and Keep Sydney Open could be in a similar position. No other parties ran in enough seats to get a good sense of the level of popular support.

The following map shows where these ‘other’ votes have come from, as a proportion of the potentially formal votes in that electorate.

It jumps out that electorates in the eastern suburbs, inner city, north shore and inner west have the highest other vote. This would partly reflect higher levels of below-the-line voting, but does also suggest that parties competing with the Greens could have done better in these seats: such as Jeremy Buckingham, Sustainable Australia and Keep Sydney Open.

In particular 24% of potentially formal votes in the seat of Sydney are ‘others’, which would point towards a good result for Keep Sydney Open (that one seat would be worth 0.05 of a quota, although not all those votes will be for KSO).

Of course it’s important to note that seats will be decided after preferences flow, not just on primary votes.

In the first two elections under the current voting system, the successful candidates for the last seats were those leading on primary votes. It was only in 2011 when Pauline Hanson was leapfrogged by Jeremy Buckingham and a Nationals candidate to miss out. Then Animal Justice overtook No Land Tax in 2015. The ‘partial quotas’ chart on Antony Green’s guide depicts these results quite well.

We don’t have a good sense of how preferences will flow. While most preferences will probably still exhaust, it’s possible to imagine an increase in preference flows following the changes to the Senate voting system.

There was a concerted effort amongst parties of the left to swap preferences. Animal Justice, Sustainable Australia and Keep Sydney Open all preferenced each other in the top three. Animal Justice and Keep Sydney Open then recommended high-ranking preferences with the Greens and Labor (which was reciprocated).

There was no such effort on the right. So this could help Labor, Animal Justice and potentially others on the left in a tight race.

So what’s the timeline from here?

The initial count of ballots (the process started on Saturday night) will continue for the rest of this week, at which point we should know how many votes are in the original nine categories started on the night. Mostly. Some postal votes will still go through this process up until Thursday 4 April.

All ballots will then be transferred to the centralised count centre where they will undergo a check count at which point we will have figures of how many votes have been counted above and below the line for every group (not just the Big Seven). Apparently we will start to get reports on these figures as of this Wednesday.

There will then be a period where we will have close-to-final initial count figures, and incomplete but more detailed check count figures. So then it will be an ongoing task to build models which analyse the detailed data and extrapolate to places yet to be completed.

And the process will finish on Friday 12 April with the pushing of the button, when we will see the preferences flow and elect 21 MLCs.

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