If the result is challenged before the Court of Disputed Returns, it will be possible to estimate the impact of the error on the Animal Justice Party by comparing the results amongst the faulty and the correct iVotes.

However, any estimate of missing votes would have to be discounted by how many below the line votes the Animal Justice Party attracted on the faulty ballots.

At the moment none of this matters. In the race for the final seat, the Coalition has 0.56 quotas towards a tenth seat while the Animal Justice Party has 0.37 quotas. The Coalition candidate is roughly 34,000 votes ahead, too many to claim that an iVote error has impacted on the result.

However, the count is not complete. Based on past trends it is clear that the Coalition total vote will decline as the count nears completion. That final count will narrow the gap between the tenth Coalition candidate Hollie Hughes and the Animal Justice candidate Mark Pearson.

Where the Count Stands now.

The Legislative Council election is for 21 members elected at large across the state. For election a party needs to fill a quota, which is set at 4.55%. For each quota filled a party wins a seat. The final seats come down to a race between partial quotas of remaining candidates. Preferences have to be distributed, but the final vacancies are likely to be filled by the candidates and parties that start the final stage of the count with the highest partial quota.

With 85.6% of the votes counted, 18 of the 21 seats are certain. The Liberal/National Coalition has 9.56 quotas and will win nine seats and is in the race for a 10th. Labor has 6.91 quotas, electing six members and certain of a seventh. The Greens have 2.10 quotas and will elect two members. Importantly the 0.10 surplus for the Greens will be distributed as preferences, and the Greens recommended a third preference for the Animal Justice Party.

By electing nine seats the Coalition will have 20 Legislative Council seats, 19 on the floor of the Council, and needing two votes to pass legislation. Compared to the previous parliament, the government will only need one small party to pass legislation rather than two.

The 19th and 20th vacancies will be filled by the Shooters and Fishers party with 0.83 quotas and the Christian Democrats with 0.61 quotas. Neither party will be passed on preferences.

Trailing them are the No Land Tax Party with 0.38 quotas and the Animal Justice Party with 0.37 quotas. These two parties are in contention with the tenth Coalition candidate for the 21st and final seat.

None of the other parties on the ballot paper have a hope of victory. If this were a Senate election, then those parties could have accumulated their votes together using ticket voting and preference harvesting to accumulate votes to win a seat. This was the method that elected the Australian Motoring Enthusiasts Party’s Ricky Muir, but this method of election from a low vote is impossible in NSW.

Under the NSW Legislative Council system, parties cannot control preferences and the only preferences that matter are those completed by voters themselves. Since the current system was introduced in 2003, 80-85% of ballot papers for excluded candidates have exhausted their preferences, which means only a party with a significant first preference vote can win election.

That is why only the tenth Liberal candidate, the Animal Justice Party, and as an outside possibility the No Land Tax Party, have any hope of winning the final seat.

Why the Last Position is in Doubt

As I mentioned above, the tenth Liberal candidate is around 34,000 votes ahead of Animal Justice, a lead that would not be under threat from a court challenge.

However, it is certain the Liberal/National vote will fall through the remainder of the count. I’ll keep a running tally of the main participants in the race over the next few days using the table below.

Legislative Council – Progress of Count – Selected Groups

	L/NP		ALP		GRN		NLT		AJP
% Counted	% Vote	Quotas	% Vote	Quotas	% Vote	Quotas	% Vote	Quotas	% Vote	Quotas
84.1	43.55	9.58	31.46	6.92	9.46	2.08	1.74	0.38	1.68	0.37
85.6	43.46	9.56	31.42	6.91	9.55	2.10	1.74	0.38	1.68	0.37
86.6	43.37	9.54	31.43	6.91	9.59	2.11	1.76	0.39	1.69	0.37
87.6	43.32	9.53	31.41	6.91	9.63	2.12	1.76	0.39	1.70	0.37
88.0	43.29	9.52	31.40	6.91	9.66	2.13	1.76	0.39	1.71	0.38
88.5	43.27	9.52	31.40	6.91	9.66	2.12	1.76	0.39	1.71	0.38
88.9	43.25	9.51	31.39	6.91	9.68	2.13	1.76	0.39	1.71	0.38
90.3	43.18	9.50	31.37	6.90	9.70	2.13	1.77	0.39	1.73	0.38

Why I say the Liberal/National vote will fall from this point is that most of the outstanding votes to be entered are Absent votes, where the Liberal/National parties do worse and the Greens better than with other forms of voting. We have already seen this trend occur as the upper house count has progressed beyond 84.0%. The final turnout in the lower house currently sits on 90.4%.

As the absent vote was added in the lower house, the Liberal/National vote fell from 47.0% to 46.7% and the Green vote rose from from 10.0% to 10.3%. If the same shifts in support were to take place between now and the end of the Legislative Council count, then the the Liberal quota total would slip to 9.49 and the Greens would rise to 2.16 quotas. The Animal Justice Party would stay on around 0.37 quotas.

If half of the Green vote had preferences for Animal Justice (a preference flow that I think is much higher than will occur), then the Animal Justice Party would receive 0.08 quotas as preferences and reach around 0.45 quotas, roughly within around 700 votes of the tenth Liberal.

The other factor that needs to be taken account of is that below the line votes are yet to be included in the count. As they are, the Coalition’s surplus beyond nine quotas will fall for two reasons.

First, the rate of below the line voting amongst Coalition voters is much lower than for smaller parties. Second, as the inclusion of the extra votes increase the quota, more of the Coalition’s total vote is soaked up by the nine filled quotas, leaving fewer votes for the surplus.

In 2011 the Coalition’s vote fell by 0.6% once below the line votes were included. If that were repeated in full in 2015, then the Coalition’s surplus quota total would fall to 0.36, behind Animal Justice.

However, in 2011 there was an extraordinary below the line vote for Pauline Hanson. Without her presence on the ballot paper, the Coalition and all other parties will do proportionally better with below the line votes in 2015.

However, both the factors listed above point to the final result being much closer than current totals suggest.

Challenging the Result

Once the result is finalised and 21 members elected, it is possible for the result to be challenged to the Court of Disputed Returns. The petition to the Court must be lodged within 40 days of the return of the writ.

The Court of Disputed Returns will be presided over by a Judge from the NSW Supreme Court.

If Animal Justice win the final seat ahead of the tenth Liberal, it is unlikely there will be any court challenge.

If the Liberals win the final seat, then clearly the Animal Justice Party and its lead candidate Mark Pearson are the likely plaintiffs. They will have to prove that an error in the conduct of the poll has affected the result.

If they prove the result was affected by error, argument would the proceed on what corrective action the Court should impose.

There are two critical tests that will apply to the question of how the result was affected.

The first is whether Animal Justice can pass No Land Tax. Animal Justice is expected to do better on preferences than No Land Tax because of the preference recommendation of the Greens. If Animal Justice does not pass No Land tax then the case has the potential to become messier and more complex as it would require estimating what would have happened had Animal Justice passed No Land Tax and then been in contention with the tenth Liberal for the final seat.

If Animal Justice passes No Land Tax, then the case would be easier, simply a case of arguing over the size of the gap between Animal Justice and the tenth Liberal and whether this was affected by the iVote error.

Why this is Different from the WA Senate case

The WA Senate case was much murkier. The issue in WA was over the order of exclusion between ninth and tenth placed candidates. Missing ballot papers prevented the Electoral Commission and the Court of Disputed Returns from determining the correct order of exclusion. Analysis of the initial count of the missing ballot papers made it clear the missing ballot papers were critical for determining the correct order of exclusion.

At the point of the disputed exclusion, three of the six vacancies had been filled. A fourth was certain to be filled whatever order of exclusion took place. But four candidates remained in contention for the final two seats, and who was elected depended entirely on the disputed order of exclusion.

Given previous court cases, the Commonwealth Court of Disputed Returns had no other choice than to void the election.

The NSW case in relation to the final Legislative Council result is potentially much simpler. It rests on a contest between two remaining candidates for the final seat. The case can proceed in a very different manner to the WA Senate case.

What the Court of Disputed Returns Might Decide.

The Court must first determine whether the error with iVote had an affect on the result.

My rough estimate is that if the result at the final run-off between the tenth Liberal and Animal Justice is under 700 votes, then there will be a case for the iVote error having affected the result.

If Animal Justice falls fewer than 700 votes behind No Land tax at a critical point, Animal Justice could argue they should have finished in front. It would be possible using the data entered votes from the election to then estimate the final race between Animal Justice and the Liberal candidate.

If the Court decides the result has been affected, it has a number of powers under Section 161 of the Parliamentary Electorates and Elections Act. It can –

(v) To declare that any person who was returned as elected was not duly elected,

(vi) To declare any candidate duly elected who was not returned as elected,

(vii) To declare any election absolutely void,

(viii) To dismiss or uphold the petition in whole or in part

In other words it is within the power of the Court to simply change the identity of the final elected member rather than order a fresh election. Given the cost and effort of holding another election, resolving the final position by simply changing the names of elected members on the writ may be the easiest solution.

Given the Coalition has already elected the nine members it needed to negotiate passage of its privatisation legislation, it would surely be in the interests of the Coalition to give up its tenth member in a close contest rather than risk an entire half-Council election of 21 members to be decided on the single issue of electricity privatisation.

There has been a previous case under proportional representation in NSW where a disputed election was resolved by changing the elected member. In the three-member Wammerawa electorate at the 1922 election, the election of WG Ashford to the final seat was challenged by JA Clark. The Parliament’s Qualifications Committee, the predecessor of the Court of Disputed Returns, ruled that the count had been conducted incorrectly, re-calculated the result and declared Clark elected in place of Ashford.

A Legislative Council challenge could not be as easily resolved as the Wammerawa case as it would be about ballots not cast rather than ballots mis-counted.

The Court would also be mindful of Section 22A(3) of the NSW Constitution that states that periodic elections for the Council should be on the same day as the general election for the lower house. Section 22A(4) then provides an out for a separate Council election if an election is declared void, but the Court of Disputed Returns would be mindful of the intent of Section 22A(3) that elections for the two houses be held together. This would be discussed if a dispute over the election of a single member could be resolved by altering the elected members rather than voiding the election of all 21 members.

In short, if the final result in the Legislative Council is close enough to challenge, and close enough that the iVote error can be ruled to have affected the result, it does not necessarily mean that all 21 seats would be forced to a new election.