The most astonishing result of the 2013 Australian election was the extra-ordinary election of Wayne Dropulich of the Australian Sports Party in the Western Australia Senate.
With a first preference vote for his party of just 0.23%, it took an astonshing sequence of exclusions and preference flows to allow Mr Dropulich to increase his vote more than fifty fold to reach the quota of 14.3%. He was elected to fill the fifth seat ahead of the Greens who had polled 9.5% of the first preference vote.
At this stage whether Mr Dropulich will take his Senate seat depends on proceedings before the Court of Disputed Returns. The Australian Electoral Commission (AEC) has petitioned the Court to void the result and order a new election for the WA Senate.
As will be outlined below, Mr Dropulich was initially defeated on the distribution of preferences performed after the initially tally of all first preferences. After the AEC permitted a re-count, Mr Dropulich was elected on the distribution of preferences that followed the completion of the second count.
However, 1,370 ballot papers could not be located for the re-count and these were not included in the second count and distribution of preferences. If the tally of the missing votes from the first count were added to the second count, the result would change again, with a critical one vote difference at a crucial point in the count resulting in Mr Dropulich failing to be elected.
Having been duly declared elected, Mr Dropulich is entitled to be sworn in as a Senator on 1 July if the Court has not decided the matter before then, or if the Court chooses to let the second count excluding the missing votes stand. If the Court orders a fresh election, Mr Dropulich would no longer be entitled to his Senate seat. If the Court does not issue orders for a fresh election until after 1 July, Mr Dropulich would be entitled to hold his seat until a judgment is given.
What was remarkable about Mr Dropulich's victory was how a party that finished 21st out of the 27 parties on the ballot paper managed to get elected. Of the 20 parties that contributed ticket votes to his quota for election, 15 polled more votes than the Australian Sports Party.
The path by which Mr Dropulich passed those 15 parties and gained their preferences was perilous. At three points in the count Mr Dropulich was the second lowest polling candidate remaining, and on each occasion he picked up preferences from the excluded candidate that allowed him remain in the count by avoiding last place.
At another critical juncture in the count Mr Dropulich was the third lowest polling candidate ahead of the Australian Christians and the Shooters and Fishers Party. Depending on the order in which the Australian Christians and Shooters and Fishers Party finished, Mr Dropulich either became the lowest polling candidate and was excluded, or he leap frogged other parties and went on to win the fifth vacancy.
The table below list the source of all votes that made up a part of Mr Dropulich's total at the point where he was elected. The table has been derived from the detailed vote count data file released by the AEC. The first 20 entries in the table are total number of ticket votes that made up Mr Dropulich's total on election. The final two entries are below the line first preferences for Mr Dropulich, and a total of all below the line votes received from multiple sources as preferences during the count. The total of votes from the Liberal Party is calculated by applying the transfer value following the election of the third Liberal Senator to the total of all Liberal ticket votes from the first count.
The '% Vote' column is the votes in the second column expressed as a percentage of the total formal vote. The '% of Quota' is the votes as a percentage of the 191,219 votes held by Mr Dropulich at the point when he passed the quota. The 'Preference number' is the preference number for Dropulich on the group ticket vote of the party that sent the votes to Dropulich as preferences. The 'Received at Count' column is the stage of the count at which Mr Dropulich received each parcel of votes.
Party | Votes | % Vote | % of Quota | Preference Number | Received at Count |
---|---|---|---|---|---|
Liberal Democrats | 44,274 | 3.4 | 23.2 | 3 | 154 |
Australian Christians | 19,676 | 1.5 | 10.3 | 11 | 146 |
Sex Party | 17,830 | 1.4 | 9.3 | 11 | 150 |
Liberal Party | 16,495 | 1.3 | 8.6 | 15 | 162 |
Help End Marijuana Prohibition | 12,740 | 1.0 | 6.7 | 3 | 150 |
Shooters and Fishers Party | 12,586 | 1.0 | 6.6 | 3 | 142 |
Animal Justice Party | 9,004 | 0.7 | 4.7 | 13 | 150 |
Family First Party | 8,303 | 0.6 | 4.3 | 9 | 126 |
Wikileaks Party | 8,129 | 0.6 | 4.3 | 3 | 130 |
Smokers Rights | 8,081 | 0.6 | 4.2 | 5 | 154 |
Australian Motoring Enthusiasts Party | 7,334 | 0.6 | 3.8 | 5 | 118 |
Australian Fishing and Lifestyle Party | 5,511 | 0.4 | 2.9 | 17 | 146 |
Australian Independents | 3,687 | 0.3 | 1.9 | 13 | 142 |
Rise Up Australia | 3,473 | 0.3 | 1.8 | 7 | 98 |
Australian Democrats | 3,266 | 0.3 | 1.7 | 17 | 154 |
Australian Sports Party | 2,866 | 0.2 | 1.5 | 1 | 1 |
Stop the Greens | 2,074 | 0.2 | 1.1 | 5 | 154 |
No Carbon Tax Climate Sceptics | 1,389 | 0.1 | 0.7 | 5 | 146 |
Australian Voice Party | 1,082 | 0.1 | 0.6 | 5 | 94 |
Stable Population Party | 1,040 | 0.1 | 0.5 | 5 | 114 |
Wayne Dropulich BTL first preferences | 108 | .. | 0.1 | 1 | 1 |
All other BTL votes received as preferences | 2,271 | 0.2 | 1.2 | .. | .. |
Total votes on election | 191,219 | .. | .. | .. | .. |
Total Formal Vote | 1,310,278 | .. | .. | .. | .. |
On the first count the Australian Sports Party had 2,997 votes or 0.23%. These consisted of 2,866 ticket votes, 108 below the line first preferences for Wayne Dropulich, and 23 below the line votes for the second candidate Al Lackovic. Lackovic was excluded at count 12 and 13 of his 23 votes flowed as preferences to Dropulich.
At the end of Count 89, only one candidate remained from each of 23 parties. During the distribution of below the line votes from the excluded candidates, Dropulich had gained 26 votes and had a total of 3,023 votes, 0.23% of the vote. At this stage he lay in 21st place of the 23 candidates on the ballot.
At the end of Count 93, Mr Dropulich was ordered at 21st position out of 22 candidates with 3,027 votes to 2,598 for the last placed Adrian Byass of the No Carbon Tax Climate Sceptics. Byass was then excluded and his total included 1,082 ticket votes from the Australian Voice Party and these had next preference for Dropulich. This allowed Dropulich to escape last place to sit 18th of 21 remaining candidates.
Last position now fell to Jane Foreman of Rise Up Australia. Her total of 3,950 votes included 3,266 ticket votes with next preference for Dropulich. This was the first of 15 bundles of votes from candidates with more votes than Dropulich. At the end of Foreman's exclusion at Count 101, Dropulich now sat in 16th place of the 20 remaining candidate.
By the end of Count 113 Dropulich had received no more ticket votes and had slipped back down the ranking to sit in 16th place of 17 remaining candidates. Dropulich trailed Richie Howlett of the Australian Motoring Enthusiasts Party by 193 votes, 7,843 to 7,649 from Dropulich. Count 114 saw salvation arrive for the second time at the last possible moment. The exclusion of the Australian Democrats released 1,040 ticket votes from the Stable Population Party with next preference for Dropulich. This put Dropulich ahead of the Australian Motorists Party for the first time and lifted Dropulich from the bottom of the pile and into second last place.
Now salvation arrived for the third time. With the Australian Motoring Enthusiasts Party now in last place and excluded, 7,334 ticket votes from the Australian Motoring Enthusiasts Party flowed to Dropulich, preferences from a party that started the count with two and a half times as many votes as Dropulich. Dropulich was now lifted to tenth place of the sixteen remaining candidates.
At Count 126 Dropulich received 8,303 Family First ticket votes and at count 130 received 8,129 ticket votes from the Wikileaks Party, two more parties that started out well ahead of Dropulich. These two bundles put Dropulich ahead of both the Australian Christians and the Shooters and Fishers Party for the first time. This was a critical result for what was to follow.
At this point, the end of Count 141, Dropulich was in eighth place of the ten remaining candidates. The two parties behind him were the Australian Christians and the Shooters and Fishers Party. If the Shooters and Fishers were in last place and excluded, Dropulich received Shooters and Fishers preferences and stayed in eigth place and then gained Australian Christian preferences at the next count. If the Australian Christians were excluded first, their preferences went to the Shooters and Fishers Party putting them into eighth place, ahead of Dropulich, resulting in Dropulich being excluded.
The count at this point was incredibly close, and the table below shows the order of the three lowest candidates and eventual outcome on the first and second counts, as well as the possible outcome if the first count of the 1,370 missing votes were included.
First Count | Second Count | Including Missing Votes | |
---|---|---|---|
Dropulich (Sports Party) | 32,829 (8th) | 32,819 (8th) | 32,819 (8th) |
Bow (Shooters and Fishers) | 23,515 (9th) | 23,514 (10th) | 23,532 (9th) |
van Burgel (Christians) | 23,501 (10th) | 23,526 (9th) | 23,531 (10th) |
Critical margin | Shooters by 14 votes | Christians by 12 votes | Shooters by 1 vote |
Last two Senators elected | Wang (Palmer United) Pratt (Labor) |
Dropulich (Sports Party) Ludlam (Green) |
Wang (Palmer United) Pratt (Labor) |
Having survived this critical count according to the accepted second count, at Count 142 Dropulich gains the ticket votes of the Australian Independents and Shooters and Fishers to pull ahead of the Australian Christians. At Count 146 he receives the Australian Christian, Fishing and Lifestyle Party and No Carbon Tax Climate Sceptics ticket votes, putting him into fourth place of the eight remaining candidates. At count 150 he receives more ticket votes from the Sex Party, Help End Marijuana Prohibition, staying in fourth place of seven remaining candidates.
Count 154 saw the exclusion of the Liberal Democrats, a party that polled more than 12 times the number of votes recorded by Dropulich. He now received the ticket votes of the Liberal Democrats, Smokers Rights, Australian Democrats and Stop the Greens to move into first place of the six remaining candidates.
Counts 158-161 exclude the National Party, putting the third and only remaining Liberal candidate over the quota. Count 162 then distributed the surplus vote of the Liberal candidate, delivering the final 16,495 ticket votes that elected Dropulich to the fifth vacancy.
This result was entirely determined by ticket votes and the chance order of exclusion in the critical race for 9th between the Australian Christians and the Shooters and Fishers Party. Of Dropulich's votes, 98.7% were ticket votes, entirely under the control of pre-arranged preference deals.
The oddity of the system is revealed if you consider what happened to the votes of anyone who voted the Australian Motoring Enthusiasts Party. Depending on the contest between the Christians and Shooters, votes for the Motoring Enthusiasts Party elected either Dropulich or Wang of the Palmer United Party. Yet those Motoring Enthusiasts Party preference played no part in determine the result, they were merely directed one way or the other depending entirely on the contest between the Christians and Shooters.
It is always a possibility that preferential voting can produce odd outcomes because of the order in which candidates are excluded. However, that possibility is made much more likely in the Senate because of ticket voting. Under non-ticket voting systems, this problem would only occur in the exclusion order of high polling candidates, not candidates at the bottom of the pile.
In systems where voters fill in their own preferences, the order that low polling candidates are excluded rarely matters, as a party with low votes tends to have more random preferences.
However, under ticket voting, a low polling candidate still has strong preference flows. In the case of WA, the contest between the Christians and Shooters produced preference flows that were tangential rather than random. The contest between two candidates with 1.75% of the vote each determined the last two seats in the contest, and both parties were already long excluded by the point in the count where the final two seats were decided.
The WA Senate count is an example of what is wrong with the current Senate electoral system. Ticket voting, which means that more than 95% of preferences are pre-determined at each exclusion, makes the order of exclusion much more important than it should be. The count becomes unstable, and whether you want to fall back on the analogy of 'butterfly wings' or 'black swans', the almost random event of the order low polling candidates finish, can have an impact that is magnified through the count.
A change to the system that required voters to enter their own preferences under an optional preferential system would give much more weight to candidates with a significant tally of first preference votes. Candidates that struggle to attract more than a tiny fraction of the vote would struggle just as much to attract more than a fraction of the preferences.
Just looking at the 'micro-parties' involved, it looks as though the Druery camp was broadly 'Libertarian' and therefore the fix was in to some extent.
I mean, unless I'm reading it wrongly, WA could just as easily have ended up with another micro-party in the 5th spot because of the preference deals Druery set up?
Am I right to understand that 'Above the Line' preferencing would deliver a result more in line with the first preferences as actually voted?
COMMENT: No, the alternative was the election of Palmer United Party as took place on the first count. Getting rid of tickets votes and allowing voters to optionally give their own preferences above the line would advantage parties that receive significant numbers of first preference votes as it would remove the ability of parties to harvest preferences.
Posted by: Megan | December 21, 2013 at 09:45 PM
The current system by which Senators are chosen clearly needs to be revisited to prevent the kind of result described above.
One simple change that could be made would be for Parties to exceed the minimum number of first preference votes in order to secure funding from the AEC - which for the Senate is 4% (for all candidates on that party's Senate ticket). Parties who fail to achieve this quota would then have all their preferences allocated immediately.
Certainly this would be a means to help ensure that micro and/or single issue parties who lack any genuine public support find themselves as representatives of the public at large. Clearly they are not "representatives" in the sense that was no doubt intended by the architects of of Constitution - nor in any common sense understanding of the term.
COMMENT: You are talking about the introduction of a threshold. The problem is at what level do you set it? Set it too low and a micro-party can still be elected by preference tickets. Set it too high and you just lock in the current players. The problem is caused by ticket voting, so I think the solution is to deal with the tickets rather than keep ticket voting and introduce an arbitrary threshold.
Posted by: Steve Robinson | December 22, 2013 at 04:48 AM
This has to be fixed. Somebody please tell us how.
COMMENT: It will be fixed, the question is by which method.
Posted by: Ruth_homes | December 22, 2013 at 10:45 AM
Antony I agree with you but I would also add that it is full preferencing coupled with ticket voting that is the problem. It is the full preferencing which causes a virtual lottery for the allocation of preferences towards the end of the count. I would suggest the NSW upper house optional preferential voting method (not count method) be examined as a fix for the Senate. NSW does not have tickets (removed after 1999 table cloth ballot paper), they allow electors to decide who to vote for above the line. Electors can enter above the line one preference or all preferences or anything in between. Typically the candidates of minor parties preferences exhaust before they can contribute to the lottery process we have seen in the Senate in Vic and WA.
Posted by: Ian Brightwell | December 22, 2013 at 03:07 PM
Antony is correct. The problem is ticket voting. My preference is for the the introduction of optional preferential voting above the line or for those voting below the line removing the requirement to vote the complete preference ticket for a vote to be valid - perhaps only requiring ten preferences.
Posted by: Doug | December 22, 2013 at 10:37 PM
Hi Antony
I think the first issue that needs addressing in the Senate Count process is the way that surpluses are distributed after the election of candidates from groups with >1 quota. In fact, that whole concept needs to be discarded altogether!
Rather than have a massive loss of preferences to other parties based on theoretical numbers, each candidate elected should be elected on; 1 - their own BTL first preference votes and 2 - the number of ticket votes required to achieve their quota from the group pool.
Both types of votes need to remain separate at the commencement of the count rather than pooling the Ticket Votes in with the BTL first preference votes of the lead candidate in each group.
Using the WA Senate vote as an example; the Liberal Party achieved the highest number of votes - 513,639, comprising 507,310 Group Ticket Votes, 4936 BTL first preference votes for David Johnston and 1393 BTL first preference votes for other candidates.
Using my process, Senator Johnston would have been elected on his own 4936 BTL first preference votes and 182,247 Group Ticket Votes removed from the Liberal's Group Ticket pool.
No Surplus Distribution required. No loss by fraction, no exhausted votes, no theoretical votes allocated to other groups to account for later in the count.
It was interesting to note Wayne Dropulich picked up 5 votes from those "Surplus Distributions". I wonder if that meant the difference between whether he ended up being in last or second last place in any of the subsequent counts.
COMMENT: And you also increase the power of party ticket votes because if you vote below the line for one of the major party candidates, your vote can never be examined for preferences. Instead all the effort you went to to mark your own preferences would be a waste of time as only your first preference would count and your vote would result in an extra ticket vote being included in the surplus.
Posted by: Jeff | December 23, 2013 at 09:07 AM
Antony Green said: "A change to the system that required voters to enter their own preferences under an optional preferential system would give much more weight to candidates with a significant tally of first preference votes. "
Antony, why do you advocate a change to an optional preferential system, instead of a compulsory preferential system (with preferences able to be entered above-the-line) ?
COMMENT: The Howard government proposed what you suggest in 2004, to abandon ticket votes and replace them with full preferential voting above the line and maintain full preferential voting below the line.
The idea was abandoned for two reasons. One, it would have put the informal vote rate through the roof. Second, it would have required a twenty fold increase in the data entry of ballot papers.
Posted by: Bosworth | December 23, 2013 at 10:36 AM
As you've said Antony, where to go from here is the issue is the question.
I proposed the following more than four years ago now in 2009, when my interest in electoral systems was more academic than practical.
http://www.dpmc.gov.au/consultation/elect_reform/strengthening_democracy/pdfs/99%20-%20Clinton%20Mead.pdf
I pointed out what you have more recently, that the options for voting in the Senate are to say the least a bit extreme. You can either vote 1, or number every box below the line. The reality is that many voters probably fall between the two, perhaps wanting to preference three or four parties.
The optional preferential system works well in the NSW LC. Counting by what is effective close to largest remainder means majors are not significantly overrepresented and minors have a reasonable chance. However 21 members are elected at the NSW LC election, not 6.
Consider optional preferential for the Senate. Unless there is a significant crash in either the Labor or Green vote, the result will be 2 Labor + 1 Green in every state. Labor+Green will have a blocking majority almost by default. The only way that could be prevented is either:
(a) Palmer beats Greens
(b) Coalition run split ticket and Nationals beat Greens
This is a consequence of the fact that under optional preferential, particularly with only 6 places, 0.7 (Greens) + 1.7 (Labor) quotas is likely to win 3 spots.
I'm not saying this to pick on Labor/Green, it could work the other way to if a significant minor like the Greens was to rise in Coalition territory. I'm just saying that OPV could mean an appropriately split two parties could control the senate with 2.4 quotas and one strong state.
The alternative is to move to D'Hondt or Sainte-Laguë (perhaps modified with preferences like the former ACT system), but it's likely to cause some cases where one side of politics gains 4 out of 6 seats in a state with a knife-edge win on perhaps 40% of the primary vote or less.
The reality is that there's around 10-15% of the vote for non-Labor/Green/Coalition parties. That's no so much of an issue in a NSWLC contest when that 10-15% is actually deciding a number of minor party spots, but with only 6 senators it's a different story. The question is what to do with that vote, much of which will expire under OPV. It's questionable whether its more distortionary to ignore it (as would be the likely reality for 70% of it without GVTs) or distribute according to GVTs.
I think we need to be careful before throwing the baby out with the bathwater. The example of microparties like Sports getting elected again is much less likely, as the cat is out of the bag. Parties will be much more careful about throwing micros up on their preference list without first thinking perhaps they'll be a threat themselves. This to a certain extent this will self correct things. I suspect the 1+%ers will be more careful about this in the future.
What I suggested in 2009, and still stand by today, is to allow optional preferential above the line, but complete the vote with the GVT of the first preference. That means 1 votes are treated the same, but if preferences are indicated the vote follows the preference of the voter and only after the voter preferences expire is the GVT followed. Whilst many people will still bullet vote 1, a significant number of particularly minor party voters will fill in a few preferences. So only 70% of ticket votes will flow, not the 98% which is currently the case. That 30% leakage will make it very hard for the real micros to get elected, as they'll effectively need 20% on the ticket vote, not the 14% as currently.
This gives the voter freedom to express their vote to the extent they choose whilst still saving the vote with the GVT. Whilst GVT preferences aren't entirely ideological, they still are significantly so, and I think one could argue that despite the politics a voter's intent is still better represented by the GVT than just expiring. Combining OPV above the line with GVTs, along with what will now be some caution by significant minors preferencing micros, I think will be enough to fix the issue without the distortion that would be caused by a 10-15% expiry of votes which may affect the balance of the senate significantly.
Data entry for such a system would be increase, but nothing that couldn't be handled. It will just be like NSWLC. The computer will handle assigning the GVTs after voter preferences have expired with ease.
COMMENT: I'm not sure why you call it optional preferential voting because at the point where the voter stops giving preferences, your system appears to default the remainder of preferences to the group ticket using a mechanism the voter would struggle understand.
You have missed another option for the Senate. Instead of having a fixed quota and a one-pass count, you can vary the quota as votes exhaust and move to an iterative method of calculating the result. As the quota declines with exhausted preferences, the count is repeated to that point in the count with the new lower quota. It avoids the problem with highest remainder systems of 4 or 5 candidates being elected on a full and fixed quota and 1 or 2 on a substantially reduced level of the vote.
Posted by: Clinton | December 23, 2013 at 01:27 PM
"You have missed another option for the Senate. Instead of having a fixed quota and a one-pass count, you can vary the quota as votes exhaust and move to an iterative method of calculating the result."
All the iterative option does is move the result closer to D'Hondt, as the reducing quota better spreads the vote among continuing candidates in the same party. The only difference from D'Hondt in the system proposed here is that parties must get at least N-1 quotas to get N seats, lest their Nth candidate be eliminated early. That's no so much an issue in the Senate.
Consider the following election, not entirely unreasonable I don't think:
45% Coalition, 32% Labor, 10% Green, 13% Other
Lets distribute this 1.5% Coalition, 1% Labor, 0.5% Green and the rest expired/other minors. Roughly a 25% preference rate.
46.5% Coalition, 33% Labor, 10.5% Green, 10% expired.
There's 90% of the vote left, so we recalculate the quota which is now 90% / 7 - 12.85%
Current candidates in the count:
4th Coalition: 7.92%
3rd Labor: 7.29%
1st Green: 10.5%
3rd Labor eliminated.
Based on the 2011 NSWLC results, about 8.6% of Labor voters preferenced the Greens, and 2% the Coaltion. Lets go round figures and be a bit generous and give the Greens 10% of Labor prefs and Libs 2%.
The elimination of the third Labor candidate is tricky because the amount of their vote we transfer is based on the final quota, which is in turned based on how much we transfer. But we can solve for that like using the following:
New Quota = (46.5% Lib + 10.5% Green + 2 * New Quota + (33% - 2 * New Quota) * 12% ) / 7 = 11.63%
That gives Labor 9.73% to distribute, which goes 0.973% Green, 0.195% Lib
Meaning the count now stands at roughly:
46.7% Colation, 11.5% Green
Based on the recalculated quota of 11.63%, Coalition is now at 4 quotas.
This goes
Coalition: 4
Labor: 2
Green: 0.
As can be seen, iteratively recalculating the quota ends up like D'Hondt near the end, favouring large parties more so than our current system. And I addressed that briefly in the original post.
A result where Labor+Green get 42% of the vote (and over three quotas after prefs) but end up with 2 seats will move the Senate toward more often having majority party control even with a minority of the vote. This is significant change to our democracy and shouldn't be taken lightly.
COMMENT: One of the consequences of optional preferential under the higher Senate quota is that it would be in the interest of parties to go to much greater effort to encourage voters to fill in preferences than is the case in the NSW egislative Council. I think the preference flows would be much higher for parties that actively campaign.
The significant change to our democracy took place in 1984 when the number of Senators per half-Senate election was increased from 5 to 6, making it much less likely that a party could achieve a majority in the Senate. If a change in the system makes it possible for a party to again get a majority of the seats in a state, then it does no more than take it back to the system that applied before 1984. Proportional representation should apply to big parties as well as small parties and permit a party that polls well to have the opportunity of achieving a majority. I'm not afraid of a party getting a majority of the vote also getting a majority of the seats.
Posted by: Clinton | December 23, 2013 at 03:17 PM
"I'm not afraid of a party getting a majority of the vote also getting a majority of the seats."
I'm not afraid of that either, but moves towards OPV, mean that parties with a minority of votes can get a majority of seats.
Currently, senate contests tend to one of three ways:
3/3 (when Liberal/National and Labor/Greens are balanced)
3/2/1 (i.e other minor) (when one of the Liberal/National or Labor/Greens pairing significantly outpolls the other).
4/2 (when one of those pairs actually gets a majority)
Any move towards OPV lessens the chance of other minors getting elected significantly, and will result in more 4/2 landslides with only a minority of the vote.
The question is what you do with the minor party vote, i.e. votes for parties 7% or under who are unlikely to get elected under OPV? If they just scatter/expire, a minority becomes an effective majority.
I wouldn't mind an increase to 14 senators per state (perhaps 13 first and then 14 over the next two half senate elections) as that would allow a majority party to get a majority of seats.
But if a voter hasn't filled in remaining preferences, I think a GVT is a reasonable way to save that vote. To not do so distorts the balance between major parties based on whether the minor parties in their space can get elected in their own right. Currently this helps Labor as the Greens can grab a spot on their own and hurts the coalition as Palmer/Shooters/Christians have would have trouble getting elected on their own in OPV. But that could change. In any case completely abolishing GVTs in a 6 seat environment can distort results and create 4/2 majorities where no majority exists.
OPV would have FPP like effects on Senate voting, increasing minorities into majorities, and lessening the role of the Senate over the long term. Despite recent success, minor parties are still under-represented based on their vote in the Senate, and this is significantly due to the small number of candidates to be elected.
COMMENT: I don't see why a ballot paper where the voter has not expressed a full sequence of preferences has to be 'saved'. If a voter doesn't have a preference for every candidate on a ballot paper, I don't see why they should have to invent such preferences to have the preferences they do have count. Nor do I think that someone else is able to infer their intent in determining a preference flow that somehow better expresses their intent than simply exhaustng.
Posted by: Clinton | December 23, 2013 at 03:59 PM
And just a comment to Megan, in 2013 the Libertarians weren't in the Druery camp at all.
Most minor parties understand the Senate voting system nowadays. And if you look at who Druery was really involved with and who got elected you'll find a bit of a mismatch.
The Druery factor is a bit overblown. Of course Druery talked to a lot of minor parties, but so did plenty of the minor parties themselves. It's just that the minor parties are busy trying to get elected, not promote a business they're running, so you might not hear about it as much.
Druery or no Druery, the minor parties were going to talk to each other, because there's 20% of Australians who don't want to be represented by Labor/Liberal/Greens. If anything, they may have done even better without that influence, because when someone is making an "alliance" which includes some/excludes others, and is getting paid by some and not by others, that tends to induce division not unity.
Posted by: Clinton | December 23, 2013 at 04:23 PM
I favour abandoning preferencing completely in favour of an open-list system of PR. It should be easy to understand that each vote counts in favour of the candidate it's cast for over all other candidates and in favour of the list that candidate is part of over all other lists. It would take away the option for votes to be transferred to a different list if the whole list failed to get enough votes to elect one Senator, but I think that's a small loss; voting patterns suggest that most voters probably don't care about that option. The counting process would be hugely simplified and the result therefore known earlier. Also, if the order of names on lists was rotated on the ballot papers, which should be relatively easy to arrange with computerised printing, then there would be no safe spots for any candidates, and parties would have a stronger motive to nominate candidates who had a personal appeal to voters.
COMMENT: Unfortunately, groups like the Proportional Representation Society, and various academics, are committed to advocating Hare-Clark with Robson Rotation, optional preferential voting and re-counts to fill casual vacancies. I don't think there is any chance of this system being introduced, which without advocacy of some other system, means there is a high probability that the major parties will just get together and tweak the system.
Hare-Clark style systems are so rare in the world of elections, and where they are used confined to electorates with relatively small voting populations, I just don't see it as a solution to Senate voting which is largely party based. There are plenty of perfectly functioning democracies with party based voting systems, but there is nil interest in such systems by electoral experts in this country because of the entrenched attitude that Hare-Clark is best.
Posted by: J-D | December 24, 2013 at 05:23 PM
The question for me is what representation do those that do not prefer the big four (Lib, Lab, Nat, Grn) get?
Whether I like it or not, if 1/6 of the people are further right than the Lib/Nats or further left than the Lab/Grns they must be entitled to some represntation.
The issue seems to be, if 1/6 are far right (or left) which party represents them and their interests?
of course, if 1/6 prefer a centerist posirion it becomes much messier.
COMMENT: If they get the vote they get elected. But if you are talking about one-sixth of the vote divided between 30 parties whose political positions extends from far left to far right, then I don't see why the electoral systems should allow entirely strategic preference deals to elect a Senator by accumulating all these votes together. Give the power over preferences back to voters, and if voters choose to vote for a party or associated party that attracts enough votes and enough preferences from voters themselves, they can get elected.
Posted by: Kevin Lawrence | December 24, 2013 at 06:27 PM
What happens if the election is done over again? Would all 6 candidates have to run again or just one vacancy has to be filled? Is there any provision for re-running a Senate election?
What would be the best reform to prevent vote harvesting from the smaller parties? Wouldn't having optional preferential voting in the Senate confuse voters voting for the House?
COMMENT: Your first two questions will be determined by the High Court sitting as the Court of Disputed Returns, but the AEC's submission argues that all six seats will have to go to a re-election. There is no provision for a Senate re-election, but if the result were voided, writs would be issued for a new election, which would mean a new close of rolls, close of nominations and polling day, as would occur if a House result was voided by the courts. Unless the High Court finds some novel constitutional reason to restrict rolls or nominations, the re-election would occur in exactly the same way as a general election.
In my view the ideal reform is to hand control of preferences back to voters which would require some form of optional preferential voting. If optional preferential voting were introduced with the current Senate ballot paper, maintaining above and below the line voting, then the confusion should be no worse than now. 98% of Senate votes are completed with a single '1'. Introducing optional preferential voting for both houses would reduce confusion and more than halve the lower house informal vote.
Posted by: Rob | December 30, 2013 at 04:17 PM
I agree that Antony makes a compelling case for changing the Senate voting system to remove the advantage given to preference harvesting.
There is a reasonable chance that Western Australians will be voting again to choose half their Senators in the next five months or so. The Federal Parliament should act soon, so the second vote will not take place under the same flawed rules.
COMMENT: That would be a course of action certain to attract a High Court challenge. The electoral act has no provisions that deal with re-elections, so some limited rules on the rolls, on nominations, could be considered to specifically cover re-elections. However, new rules for counting at re-elections would be very difficult to justify unless the High Court orders a partial re-election.
Posted by: robd | December 31, 2013 at 05:50 AM
There is an ever-present logical fallacy that keeps popping up to annoy me here, and that is that anyone who voted for any candidate outside the top 4 (LIB, ALP, GRN, NAT) has voted AGAINST the top 4 rather than FOR a specific candidate, and should therefore be satisfied with the election of any other candidate. By that logic, if 14.3% of the electorate scatter their votes among 20 or 30 minor parties then democracy is served by electing any single one of them.
Should we believe that the 1-2% of voters who consistently support the Australian Christians (or Christian Democratic Party before them) would be perfectly happy to be represented by a senator from the Australian Sex Party or Help End Marijuana Prohibition, purely on the basis that they didn't vote for one of the top 4 and the elected senator wasn't in the top 4?
In short, no. That is simply not how any electoral system in the world should work.
I would love to see the Sainte-Lague method implemented for Senate elections across Australia. I expect this would have delivered 18 LIB (or LIB/NAT), 14 ALP, 4 GRN, 2 XEN, 1 PUP and 1 LDP (name confusion and prime positioning can be an argument for another day) - a very reasonable cross-section given the changes in primary vote in each state. I have no problem with PUP winning a seat in QLD or XEN winning 2 in SA as that's what the primary votes actually look like - completely defendable without resorting to weird logic.
COMMENT: As I keep pointing out, why should Mr Dropulich have been elected ahead of 14 other parties whose preferences he received but who received more votes than him. The level of first preferences a party receives is based on their level of support and level of campaigning, yet the preferences they receive depends not on the views of voters but on the arranged preference deals.
Posted by: Scott | December 31, 2013 at 11:15 AM
Article could have been improved by defining "ticket vote". Most of us know what it means but many may not.
If you really want radical, each half senate should be appointed by random allocation of citizens from each State, like jury duty. It would be more representative than the current system. You serve your term with all expenses paid.
COMMENT: I think it might be a hard concept to sell at the constitutional referendum that would be required for its implementation.
Posted by: Kevin McCready | January 01, 2014 at 11:51 AM