This is Lucas's Typepad Profile.
Join Typepad and start following Lucas's activity
Join Now!
Already a member? Sign In
Lucas
Recent Activity
Antony, I know that in the Senate of 2011-2014 Labor and The Greens had a combined total of 40 Senators i.e. 31 Labor and 9 Greens (and I'll again point out for the record that I'm a member of The Greens). I realise that in Tasmania, at both the 2007 and 2010 half Senate elections, It was 3 X Labor 2 X Liberal and 1 X Green that got elected and that had to be the outcome in at least one of those half Senate elections in order for The Greens to be the sole balance of power party in The Senate during the 2011-14 period and hence was pivotal for a lot of The Gillard Government's legislation passing. My question is, am I correct in presuming that had the new Senate voting system been in place at the 2007 and 2010 elections, that you would still have had the same outcome in Tasmania (Hence putting the nay sayers, who claim it will end up leading to obstructionist Coalition blocking majorities, in their place)? COMMENT: Most probably.
1 reply
Antony apologies if you've already answered this question in a previous blog post, but I'm curious as to whether transfer values will be weighted under the new system, as not every voter will allocate a preference for candidate(s) who remain in the count when surplus votes are distributed. I'll use a basic hypothetical example. Let's in Victoria a group get's 1.2 quotas on primary votes and exactly half of the ballot papers for the group in question have just been marked 1 ATL with no further preferences indicated and the other half have at least six preferences indicated with all of those being allocated to candidates who remain in the count when the 0.2 of a quota is distributed as surplus votes. Would the transfer value of the ballot papers that have expressed preferences be double of what they would be if every ballot paper making up those 1.2 quotas expressed preferences to candidates remaining in the count. Or is the new system not going to be weighted? COMMENT: Say a party's ticket has 1.2 quotas, assume all the votes are at the top of the ticket and trickling down through the candidates, and assume half of the party's voters only number candidates in the ticket and have no preferences outside of the ticket. The top candidate would be elected with a quota and the 0.2 surplus trickle to the second candidate. This would be all 1.2 quotas of ballot papers at Transfer Value (0.2/1.2 = 0.1666), which corresponds to 0.2 quotas of votes. (Remember that in PR-STV systems, the number of votes is equal to the number of ballot papers times the transfer value.) If that 0.2 is excluded later in the count and distributed as preferences, then with 50% of the ballot papers having no further preferences beyond the ticket, then 0.6 quotas of ballot papers have preferences and 0.6 have no further preferences. So that is 0.1 quotas of votes with preferences and 0.1 without. So when the 0.2 quota of votes is distributed, 0.1 are distributed and 0.1 exhaust. This is how the current formulas work, and at this stage there is no change to the formula. What could happen is that exhausted votes could be excluded before calculating transfer value, as long as the transfer value of a ballot paper is not allowed to increase in the formula. That would actually have no practical difference in the above example as the transfer value would have been calculated early in the count when the ballot papers did not have exhausted preferences. There are ways that you could squeeze all the votes with preferences to the end of the ticket, so that the full 0.2 surplus of votes could be distributed. But that would be changing the system to give more weight to the preferences of groups with more than a quota compared to groups with less than a quota.
1 reply
I'm all for the new 1-6 ATL rule and I'm all for bringing BTL voting into line with the Victorian upper house and hence having the 1-6 option for BTL as well as ATL and hopefully that's an amendment that either Lee Rhiannon and/or Nick Xenophon successfully get through. Ideally, I'd like to see fixed four year terms implemented for both the HoR and the Senate, so that every four years, the Senate quota is 7.69% at every election instead of DD elections only. I'm all for a happy medium of getting rid of preference harvesting, but at the same time, not making entry to the Senate too difficult for minor parties and independents and the 1-6 rule combined with a constant 7.69% quota would achieve exactly that. Antony, I take it implementing fixed terms of that nature at a Federal level would have some Constitutional barriers? COMMENT: Any change to the Senate's term requires a referendum. In 1988 a referendum was put for a four year term for the House and the Senate. It brought all Senators into a single cycle, electing all Senators each House election, but did not include fixed terms. The proposal received only 32.9% support.
1 reply
Antony, as a Victorian, I decided to give the mock version of NSW's iVote a go. I discovered that the system would not allow for preferences to be marked both above and below the line on the same vote. This makes me curious about the proposed changes. I'll use the below example (there won't be so few candidates in the States, but maybe in the Territories): There's 8 groups with 2 candidates per group and The Greens are one of them (and for what it's worth, I declare that yes I am a Greens member, but views expressed are my own). All 8 have ATL voting squares, but there are 3 ungrouped candidates who are all left of centre. This makes for a total of 19 candidates. Using NSW rules on a paper ballot, if I voted 1 Greens ATL, followed by 2, 3, 4 BTL for the 3 ungrouped candidates and finally 5, 6, 7, 8, 9, 10, 11 ABL in all of the remaining ATL squares, would this be A: considered to be a formal vote and if so, B: considered to be a full preference allocation for all 19 candidates? Sure its near impossible for an ungrouped candidate to be electable in a system with ATL voting squares (in which case I'd simply mark every ATL square and leave every BTL square blank if in doubt), but I see no reason for what I suggested to be deemed informal since the intention is clear. Finally, with regards to not making the counting process too cumbersome for AEC, would a reasonable interim measure be to adopt the NSW rules in the short term to immediately get rid of preference harvesting, but have a 10-15 year plan of implementing sophisticated OCR scanning technology, following which, would then allow for ballot papers to instruct more than one preference ATL so that exhausted votes will minimised without creating massive headaches for AEC staff? COMMENT: It is perfectly sensible that you cannot vote both above and below the line when using iVote. This is allowed as a savings provision, not as a valid method of voting. The rule is there because while the ballot paper itself is informal, part of the ballot paper either below the line or above the line might be formal on its own, and therefore the voter's expression of preferences where possible to be determined should be admitted to the count. The savings provision simply determines which part of the ballot paper is checked for formality first. Your example produces an informal BTL vote as there is no first preference below the line, only above the line. You can't mix and match the votes as your '1' above the line has preferences for that group's candidates below the line so you get multiple '2', '3' preferences if you try and mix and match. The voter's intent is not clear on a ballot marked above and below the line, so the savings provision says if an intent is clear BTL, then count it, and if not see if the voter's intent above the line is clear. Your example ballot would be admitted to the count by the single '1' above the line preference. The BTL is informal. Not OCR, computer voting. The NSW iVote experience has shown that voters will fill in more preferences when voting on-line. However, the basic problem is whether voters have preferences or not. In the end it is how-to-vote cards that encourage voters to give preferences. As we know from Tasmanian elections where there are no how-to-votes, you get reasonably preferences flows between known candidates, but otherwise preferences tend to be random. In the lower house 80-85% of ballot papers do not require any preference other than the first preference. Preferences for the two leading candidates in most counts are completed for formality purposes.
1 reply
As I read on The Age recently: "Clive Palmer has moved to lock in a powerful preference deal with two crossbenchers as the mining magnate weighs up a dramatic switch to the Senate at the next election". Antony, in light of this, even if Senate voting reform is passed through parliament before this years election, will it be safe to presume that it won't implemented in time for the election? Given the electoral commission was quite stringent in deregistering parties that it did not believe had 500 members (such as Australian Sex Party, Australian Democrats and the DLP to name a few), could this have been the electoral commission's contingency in case Senate voting reform can't be implemented in time for this years election? Or will all the other micro parties that have registered recently mean we'll likely once again see more than 100 candidates in Victoria and/or New South Wales and have six point font size on the ballot papers yet again? It seems Clive Palmer has lack of Senate voting reform to thank for his chance of reelection to Parliament hanging by a thread, whereas it would be absolute zero if Senate reform was passed, be it the form of what Bob Brown in 2010 and more recently Nick Xenophon recommended (i.e. a minimum of three preferences above the line) or the form of the NSW upper house (i.e. a single first preference above the line only counting for that group). COMMENT: If something isn't done we will see ballot papers with more candidates and smaller fonts than in 2013. I don't think Xenophon's idea of three preferences above the line will be accepted. Anything other than the current option where the majority of votes are a simple '1' above the line will be very difficult to implement. I expect that if the government doesn't do anything to reform the ticket voting system, then at the last minute there will be a massive increase in deposit fees, most likely a big fee to lodge a group ticket vote. The AEC has enormous and justified concerns about the size of the ballot paper interfering with the conduct and outcome of the election.
1 reply
Antony, In light of the fact that under the current rules, a Senate group must have a minimum of two candidates for an above the line voting square, then would an ideal happy-medium be to introduce above the line preferential voting, but only require a voter to mark a minimum 3 boxes above the line (or a minimum 6 boxes below the line) for her/his vote to be formal (with further preferences optional)? In the Senate and the Victorian Legislative Council, you have the issue of preference harvesting (and having the option of only marking 1 to 5 below the line in the latter, seems to have had little impact on minimising preference harvesting), but in the NSW Legislative Council, you have the issue of exhausted votes. Having the option of only voting 1, 2, 3 above the line or 1, 2, 3, 4, 5, 6 below the line would eliminate preference harvesting, but at the same time would result in a lower amount of exhausted votes in that many micro party voters, who might have otherwise just voted 1, would likely give their 2nd and/or 3rd preference to a group that would remain in the count. Sure there would still be exhausted votes from those who vote 1, 2, 3 above the line for three micro parties, but at a guess, the proportion of exhausted votes would still be substantially lower than NSW Legislative Council elections (and as long as AEC staff stress to voters when handing out ballot papers, that a minimum of 1,2,3 above the line is required, I'd imagine there wouldn't be too many informal votes as a result of said reform). COMMENT: Yes, but I wouldn't obsess about the number of preferences. I'd adopt the ACT procedures where the ballot paper instructs that five preferences be filled, but formality requires only a single first preference. You could do that in the Senate because surplus to quota votes can be weighted out in the transfer formulas. You can adopt ballot paper instructions that try and increase the number of preferences without pushing up the number of informal votes.
1 reply
Lucas is now following The Typepad Team
Jun 13, 2015