JSCEM Report 2013 Australian Federal Election

JSCEM Final report of the 2013 Federal Election has been published.

http://www.aph.gov.au/Parliamentary_Business/Committees/Joint/Electoral_Matters/2013_General_Election/Final_Report

Missing is the much needed review of the method of counting the senate election.

The need for

1. A Weighted Surplus Transfer system.
2. A reiterative system of distribution of preferences from excluded candidates where the count is reset and restarted on each candidate(s) exclusion
3. The scraping of the Droop quota and the adoption of a pure proportional representative model (x/y)
4. The retention of Above-the-line (ATL) Group voting with the allocation of ATL voting to be equally distributed to each candidate within the group. The group maintaining the right to determine the order of exclusion within the group.

The guiding principle being

“That votes from excluded candidates should be redistributed and counted as if the excluded candidate(s) had not stood

The current system is disproportionate and favours minor parties/Candidates and does not reflect the voters intentions.

A full value vote should always be allocated to the first available continuing candidate without skipping and jumping.

A reiterative counting system would also allow of the adjustment of the quota for election on each iteration, facilitating optional preferential voting..

The current outdated system of counting the vote was designed to facilitate a manual count.

With the use of computer based technology there is no excuse or justification to maintain the outdated flawed system of counting the Senate vote.

Preliminary Analysis shows that the ALP fell short by one percent of the quota.

In a good example that demonstrates the distortion in the way the VEC counts the vote Preliminary analysis of Western Metropolitan vote shows that had a more accurate and proportional counting system been used Greens candidate Colleen Hartland would have won the seat with three votes above quota and the ALP falling short by one percent of quote just over 700 votes.

This is a stark contrast with the method used by the Victorian Electoral Commission as outlined in the legislation which at best can only be described as semi-proportional.

In our analysis we used the Wright system which included a weighted value method to calculating the surplus transfer value and on every exclusion the ballot was reset and the count restarted and votes redistributed as if the excluded candidates had not nominated. The quota for election was adjusted to accommodate votes that existed throughout the count. The Weight system uses a single transaction without segmentation and only surplus votes are distributed in a single iteration of the count. The process of iteration continues until all vacant positions are elected. ION analysis of Western Metro the vote progressed until there were only six candidates remaining in the final iteration count.

Similar results were shown using the Meek method of counting.

what this demonstrates yet again the need for the state government to abandon the outdated distorted and accurate system currently in use. Had the number of voters not exhausted their ballot papers or Victoria had retained a full preferential voting system the outcome of the election may have been different.

In 2007 the Queensland Senate election which uses the same flawed system of segmentation as exists in Victoria Greens candidate Larissa Waters missed out in being elected. Had a reiterative counting system been used instead Larissa Waters would have been elected.

The current system was designed to facilitate a manual count and with the use of computer based technology is no longer warranted or justified.

Wright System Reiterative Count - Value based Surplus Transfer - Single transaction per candidate Final Iteration  422588 Ballot papers minus 1205 ballots which exhausted with-out value expressing no preference for any continuing candidate
ID	Candidate	Group	01-001 Primary	01-002 Surplus PAKULA, Martin	01-003 Surplus FINN, Bernie	01-004 Surplus EIDEH, Khalil	01-005 Surplus ELSBURY, Andrew	01-006 Elected HARTLAND, Colleen
3	PAKULA, Martin	ALP	196632	70229	70229	70229	70229	70229
4	EIDEH, Khalil	ALP	1242	126653	126653	70229	70229	70229
5	SMITH, Bob	ALP	1411	1736	1761	57929	69419	69419
15	HARTLAND, Colleen	GREENS	69582	69987	70124	70181	70237	70237
20	FINN, Bernie	LIBERAL	151299	151299	70229	70229	70229	70229
21	ELSBURY, Andrew	LIBERAL	1217	1250	82015	82015	70229	70229
	Sum		421383	421154	421011	420812	420572	420572
	Quota		70229
	Exhausted	with value		229	372	571	811	811

Andrew Bartlett statements of misconception

Andrew Bartlet a disgraced former Australian Democrat Senator come Greens failed candidate, continues to demonstrates his lack of understanding of the Australian electoral system and the limitations of ABC Antony Green's calculator. A numbers man he is not.  If you follow the Bludger Tea Leaf chat fest then the one to watch and take seriously is Kevin Bonham, he has the best grasp of all. (Yours truly Democracy AT work was banned from participating for the during of the campaign because of offending the Dr by publishing anti Greens statements and highlighting the flaws in the system - Flaws that may very well effect the outcome of the last and only seat on contention. Western Metro - Even though the campaign is over the "good" Dr has still continued his censorship).


On the Dr Bludger Green Tea web site Bartlett made a few comments that need exposure and clarification.

Andrew Bartlett

Posted Tuesday, December 7, 2010 at 3:00 pm | Permalink

I see with 72.39% of the vote in North Metro now counted, Stephen Mayne has now moved ahead of the Greens (using the Antony Green calculator which is ATL only of course) by 341 votes, which then easily cascades him to eventual victory – which would hold the Lib/Nats to 20 seats and give him and the 2 or 3 Greens balance of power.

After 91.2% of the vote in West Metro, Colleen Hartland from the Greens has a quarter of a toenail over the line for a quota (by about 45 votes).

If they stay that close then BTL and exhausts will decide the final result – given the larger number of candidates and exclusion points in North Metro that one is a bit harder to pick, (and with a sizable vote still to count, I don’t think Labor could be ruled out for that one yet either). If I had to punt I’d go with the Libs (although I’d personally I’d prefer Stephen Mayne)

West Metro looks so close that I guess we’ll just have to see what pops out when the button is finally pushed.

Stephen Mayne has been out of contention for days, certainly well before December 7 when Bartlett published his comments. Bartlett from Queensland, has not kept up with the facts instead he relies on Antony Green's flawed  and limited calculator for his predictions. North Metro is/was not close and the Liberals will win the firth position by a margin of 10,000 or more votes.  Antony Green in developing his calculator makes the assumption that every vote for a group candidate is a a ticket vote and he applies the distribution of preferences as such. (This is a reasonable assumption in a full preferential ballot and early on on the count but it comes unstuck and unravels quickly in a close ballot).  Only a fool would take it into consideration in a close contest.  One issue of contention with Green's calculator is that he did not take into consideration Adrian Whiteheads 347 BTL preferences,  He should have included him in his calculations and distributed his preferences according to his ticket. If you make an assumption for Party BTL preferences you should do likewise for the independents also.   At one stage in the count Whiteheads preferences where crucial to ousting Stephen Mayne from the count but Greens calculator did not pick this up as he had excluded him from the count without taking into consideration the fact that he preferences the Greens then exhausted.  Whiteheads 347 odd votes put the Greens ahead of Mayne much early then Antony Green's calculator had indicated.

Andrew Bartlett

Posted Tuesday, December 7, 2010 at 5:59 pm | Permalink

dgh1

I’m agreeing with (or relying on) Antony (or his Upper House calculators anyway).

Thanks – I hadn’t noticed the overall statewide increase – from memory that’s up about 0.6% from election night. The Green vote often goes up a bit with absentee and to a lesser extent pre-poll votes, although is usually lower with postals. I haven’t been following the VEC site closely enough to see whether there are mainly absentees or postals left in the Upper House counts.

Increased primary vote from the remaining count is the main thing the Greens need in West Metro, as they’re only getting preferences from the Sex Party and need some buffer for BTLs. (although ironically if the Green vote goes up much in Metro North, it will mean the seat probably ends up with the Libs rather than Stephen Mayne).

In another irony, after being put dead last by the Libs everywhere, it looks like Greens preferences will be crucial in ensuring the Libs win the final seat in Northern Victoria region ahead of the Country Alliance.

The fact is in Northern Victoria the Greens vote  is used to inflate the ALP ticket vote which kept Country Alliance in the race much longer then they should have been due to a flaw that Andrew Bartlett denies exists. Even Antony Green has acknowledged this flaw when analysing the 2007 Victorian Senate and 2010 Queensland Senate vote.  In Northern Victoria Reguion the Greens vote drops in value the same extent that the ALP ticket vote increases in value thanks to the flaw used in calculating the surplus Transfer value. As it turned out the Liberal party vote increased as the count progressed, outweighing the distortion in the method of calculating the Surplus Transfer value.  

The other aspect that Bartlett still does not understand is that the order and method of distribution of excluded candidates does affect the outcome of close elections.  In 2007, in Bartlett's home state of Queensland,  Greens Senate candidate, Larissa Waters, was denied a Senate seat as a result of the flawed distribution system where votes are distributed from the bottom of the deck skipping elected candidates and landing on continuing candidates at a higher value then they should be.  If you recount the Queensland 2007 vote excluding all candidates except the last seven  standing Larissa Waters would have been elected.  This simple exercise demonstrates the extent of the flawed system of Segmentation of the vote. The system of segmentation is outdated and should be abolished -with the use of computerised counting system we no longer need to maintain this illogical and unfait system. Something that Bartlett does not understand


All indications are that the method of Segmentation will play a significant role in the outcome of Western Metro, where the results of the election are very much decided by below the line ballot papers and the impact of Victoria's new Optional Preferential rules. Another compelling reason why Victoria should adopt a reiterative counting system, In a reiterative counting system on each exclusion of a cnciate from the count the ballot is reset and restarted, redistributing all preferences as if the excluded candidate has not stood. With a weighted Surplus Transfer value and the use of computer technology there is no need for Segmentation as all votes can be distributed in a single transaction.  The process of reiteration continues until all vacant positions are filled in a single iteration following the distribution of any Canidate's surplus votes. The same process and rules would also apply to single member electorates.

Under a reiterative counting system the quota is adjusted after the primary distribution, taking into account any ballot papers that may exhaust.  If a vote for a minor party then preferences a major party that vote would be transferred to the candidate of the voters choice and form part of that candidates surplus which would later be transferred at a lower value. Under the current rules the vote is applied to a lesser preferred candidate at a higher value.  In a fair and equitable counting system each vote should be treated equally and distributed according to the voter's intention..  Either Meek or Wright fulfil these principles.  Meek being a non-linear counting system where as Wright is refinement of the existing Senate rule and is liner in process.

Andrew Bartlett befroe losing his seat was a member of the Parliamentary "Joint Select Committee for Electoral Matters" (JSCEM) and he still does not understand how the system, with its inbuilt flaws, works. Too much drink and not enough research.

Western Metro and the Victoria's Optional Preferential vote

One of the unknown factors in the Western Metro ballot is the effect of the optional preferential ballot.  Under new rules introduced in this election is that for a below the line vote to be valid the voter must number one to five candidates of their choice.

There is no logic that justifies a five minimum Optional Preference but it makes the politicians that introduced it feel good. In fact it may end up working against them.

All parties, except Sex Party and Family First, have nominated a full team of five candidates. (None would ever elect all five - they just fill up the ballot paper and minor parties make a contribution towards consolidated revenue in lost deposits) the Optional Preferential rules means that voters can just preference their group and then exhaust.  This has the likely impact of working against the ALP as Family First voters are more likely to exhaust after preferencing the DLP or Liberal Party.

DLP and Liberal Party can exhaust within their own group and the DLP vote will not travel beyond.

In the Liberal Party's case below-the-line votes picked up in the count that do exhausts with the Liberals number two candidate will be picked up in part and carried on by the Liberal Ticket vote which increases in value at the expense of any non-Ticket vote it receives along the way. This is due to the formula used to calculate the Surplus Transfer value which increases the value of the party Ticket vote disproportionately as it moves forward. (Similar inflation of the ALP and Liberal Party tickets also occur in the election of the Parties previous candidates. In most cases it will go unnoticed as most votes travel in the same direction but the increase is there and can provide a winning margin in a close election as we are potentially seeing unfold in Western Metro)

Sex Party, as with Family First, must preference beyond their own Party in order for their vote to be considered valid. They could preference three of the already excluded or elected candidates in the race at the time Sex Party's vote is distributed, in which case the vote will exhaust carrying its value with it.  An exhausted Sex Party vote reduces the Greens chances of reaching quota in the same way that a DLP or FF exhausted vote reduces the ALPs overall vote. If a Sex Party vote lands on Labors third Candidate or the Liberal party second candidate it will move Labors chances forward or if it land on the Greens add to the Greens total,.

The Greens need 90% or more Sex Party's below the line preferences to remain in the contest

The unknown question is how many below-the-line votes will exhaust and will it play a decisive role in who is elected?if this is the case then the system will take over and not the voter and as explained earlier teh advantgae is with the major Ticket votes. If the Greens are elected in fourth place then it is all academic.

The better alternative

in order to limit the system providing an advantage to one group or anopther the best option is to adopt a weighted Surplus trabfer value foirmula and do way with segmentation and have one transaction per candidate.

In a reiterative count, such as applies in the Wright System, the ballot is reset and restarted on every exclusion. the quota is recalculated following the distribution of the first available preference. It avoids the need for segmentation of the vote.  An optional preferential vote could land on one of the previously declared elected candidates and form part of their surplus if it has no place to go it would just fall off the table and be considered in the same was as an informal ballot. Overall a reiterative count best reflects the voters intentions in the preference nominated.  It treats each ballot paper equally and when a candidate is excluded form the ballot the vote is distributed as if the excluded candidate had not stood.

In 2007 the distortion in the segmentation of the vote denied the Greens candidate, Larissa Waters, a Senate seat. (If you recount the 2007 Queensland Senate count excluding all candidates except the last seven standing (Three ALP, 3 LNP and 1 GRN) Larissa Waters should have been elected.

Flaw in counting system to deny Liberal Party representation in Northern Victoria

A flaw in the way the Victorian Electoral Commission calculates the Surplus transfer value may deny the Liberal/National Parties a third seat in Northern Victoria.  ALP ticket vote increases disportionaly in value at the expense of the Greens  delivering a bonus 989 votes to Country Alliance allowing them to win the last seat not on merit but on the back of a flaw in the system.

UPDATE:  Latest counting (@ 90.76% counted) has the Liberals move ahead of Country Alliance but the flaw in the counting system still remains although the extent of the distortion doees not effect teh outcime of the overall result.  Liberal party win last seat.

Based on the data extracted from ABC Antony Greens Election calculator at 68.68% of the vote counted (quota 47735)

The ALP ticket vote increases in value as a result of the formula used to calculate the Surplus Transfer value.  (Surplus divided by the number of ballot papers)

In calculating the surplus transfer value the VEC divided the surplus equally by the number of ballot papers as opposed to taking into consideration the weighted value of each vote.

At the point where the ALP number 2 Candidate is elected (Count 9)

Vote source	Papers	Value	Sum	Percentage
ALP Ticket	77912	0.387322	30177	57.40%
Greens Primary	22393	1.000000	22393	42.60%
Total Votes	100305		52570	100.00%
Less Quota			-47735
Surplus			4835

Weighted Value of Surplus

Value of ALP Ticket	77912	0.035623	2775	57.40%
Value of Greens Vote	22393	0.091973	2060	42.60%
Total Surplus	100305		4835	100.00%

Value allocated by VEC

Value of ALP Ticket	77912	0.048203	3756	77.68%
Value of Greens vote	22393	0.048203	1079	22.32%
Total Surplus	100305		4835	100.00%

The Greens vote has been devalued 989 votes. The ALP ticket vote increased in value by the same amount.  This is greater then the winning margin allocated to Country Alliance

Solution: Introduce a fully proportional counting system such as the Wright system or the Meek method.
One vote one value - Single transaction per candidate (Surplus or exclusion).Democratic Pure simple the Wright method

Antony Green's confusing flaw explanation explained in more detail by example

Antony Green in explanation of the Senate and Victorian upper-house counting rules tried to outline the flaw in the calculation of the Surplus Transfer value used to transfer a candidate's excess value of votes.

In presenting his arguments Green uses quotas, which to some may be confusing and difficult to understand.

The Senate and Victorian formula used to calculate the Surplus Transfer Value is based on the number of ballot papers not the value of the vote and as a result this distorts the proportionality of the election count. In short the system used is not accurate. It was put in place to facilitate a manual counting of the Senate vote, last century, when they did not have computers to assist in the counting of the vote.

To try and limit the impact of this flaw they introduce segmentation - the breaking down of the vote count into parcels or bundles of votes that held a common value.   In Tasmania and the ACT they only use the last bundle of votes received to distribute a candidate's surplus. (This has another effect that is equally flawed in principle and its execution - details not covered in this discussion)

The use of a segmented count is akin to dealing from the bottom of the deck in a game of cards. It in itself distorts the outcome of the election as was the case in 2007 Queensland Senate count where Green candidate Larissa Waters was denied representation as a result of this flaw in the counting process.

Western Australia State Government legislated to correct the flaw in the Senate Surplus Transfer Value but they maintained the method of segmented distribution of excluded candidate votes.

With the use of computer aided counting systems there is no need or justification to retain the flawed surplus transfer value calculation and with a value based formula there is no need to retain segmentation either.  All votes should be transferred in a single transaction - one transaction per candidate.  Pure simple and proportional

Senate and Victorian Legislative Council formula

Surplus transfer value (Stv) equals Candidates surplus value (Csv) divided by the Candidate's total value of votes (Ctv) divided by the total number of ballot papers (Tbp).

The problem with the formula is that they use the number of ballot papers as the divisor not the value of the vote at the time it is being distributed . This problem is highlighted when a candidate is elected in  a deferred count where the value attributed to each ballot paper is significantly different. The impact of this flaw in the system is it inflated the value of the major party ticket vote and devalue the independent below-the-line Minor Party vote.

The impact of the distorion in the count can be explained by looking more closely at the 2010 NSW Senate count and the distribution of the Liberal National Parties third candidate surplus votes using real numbers not percentage of quotas as in the Green explanation. In NSW the distortion delivered the LNP ticket a bonus value of  14,317 equivalent votes which is enough in a close election to change the result

In 2007 Victorian Senate count  the distortion in the calculation of the Surplus transfer value came close to defeating ALP's Senator David Feeney as the LNP ticket gained a "bonus" of over 7,000 votes which was then transferred to the Greens candidate. The 7000 votes effectively "stolen" from Family First, the DLP and one Nation all of which opposed the Greens candidature.

If we counted money and dividends as we count votes our financial system would collapse over night

Two wrongs do not make a right

In the continuing debate over ABC Electoral Analyst, Antony Green, censorship of public debate we present another example of Green censorship and his imposition of central control to avoiid criticism and disclosure


On Anthony Green's censored ABC blog site Antony Green entertains the debate about segmentation and responds leaving out crucial details and facts

"If all votes were distributed at once, a candidate could end up with a sizable surplus and from ballot papers at different values, and this would make the distortion from the transfer value calculation even worse. Breaking the votes into smaller bundles attempts to limit the distortion, as does doing the distribution in order of descending transfer value."

Two wrongs do not make a right.

The method of segmentation was devised to facilitate a manual count and minimise the errors that exist in the method used to calculate the surplus transfer value. There is no logic to support it, with computer aided counting there is no longer justification as a computerised count can be performed within ten minutes to three hours depending in the number candidates and number of iterations required

The Western Australia system only fixes the flaw that exists in the calculation of the Surplus Transfer Value. It does not address the distortion that exist in the segmentation of excluded candidate votes which has been left in place.  The distorion in segmenation of the vote resulted in thew rong p[erson being elcted in the Queenland 2007 Senate election. In Tasmania and teh ACT they only distribute the "last bundle" of segmented votes which is even worst as it goves more wieght to one segement of votes then it does to other segments, (IE teh vote is not distributed equally).   Both the Senate and Tasmainain/ACT systems distort the proportionality of the count and as such the results of the election.

THE WRIGHT SYSTEM

==============
The Wright system uses a reiterative counting process which seeks to address both of the idenitaled issues with segemnatuon and the flawed caklculation of the surplus transfer value, including issues related to exhausted ballots (Votes that do not express a valid preference for a continuing candidate).

The Wright system distributes only primary votes and surplus distributions in a single iteration. If all position are not filled in a single iteration then the candidate with the lest number of votes is excluded from the count and the count is reset and restarted and preferences votes reallocated as if that candidate had not stood.

The quota is recalcualted following the re-distribution of the primary vote. Any candidates that have a surplus of votes their surplus is redistributed proportionally to the value of the surplus and the value of each ballot paper, using what is referred to as the "weighted inclusive gregory transfer method"- makes some people feel good to use such titles)

The surplus vote is weighted and distributed based on the value of the vote not the number of ballot papers (As is the case in the Senate and Victorian Upper house counts).

The process of iteration continues for each exclusion until all vacant positions are filled.

This process outline in the Wright system removes the flaws and distortion that has unnecessarily been built into the Australain voting system over the years.

It does make a difference.

In the 2010 NSW Senate election the LNP group ticket vote increased in value disproportionately by over 14,000 votes as a result of the flaw in the way the Surplus Transfer value is calculated

In 2007 Victorian Senate count the LNP group vote increased in value by over 7,000 votes which could have resulted in the ALPs David Feeney losing out to the Greens who received the bonus 7.000 votes at the expense of One Nation. Family First and the DLP, all who did not support the Greens candidature.

In 2007 Queensland Senate Election Larissa Waters was not elected to office because of the method of segmentation. If you recount the 2007 QLD senate vote excluding all candidates but the last seven standing (3 ALP, 3 LNP and 1 Grn) Larissa Waters whould have been elected.

This of course suits the main parties who are sometimes the benefactor of the flawed counting sytem but it comes at the cost of devaluing  minor party votes.

If we counted money and allocated dividends, as we count Senate votes, our monetary system would collapse overnight. We argue in court about single member states that are won or lost by less then 10 votes but we blissfully ignore the upper-house system which can be lost by a distortion in the count that represents thousands of votes. A distortion that should not exist and that we must correct.

A fair, accurate voting system is not too much to ask for is it?

Green breached ABC charter

ABC Electoral Analyst, Anthony Green, continues to edit out and censor discussion and debate on our electoral system, leaving him and then ABC open to allegations of bias and failing to fulfil the ABC charter and responsibilty to provide free open discussion of public interest.

More examples of information that Antony Green refuses to publish. All factual and all of public interest

Segmentation was designed to facilitate a manual count.

There is no rational or logic supporting the continuation of segmentation of the vote.

If a candidate is excluded from the count then their votes should be redistributed as of that candidate had not stood.

This is best achieved by restarting the count and redistributing all the votes on teh excludion of any candidate from the count.

The only distribution should be primary votes and surplus distributions.

If all position are not filled in a single oetraton of the count then the candidate with the lowest vote should be excluded and the count reset and recounted. (the Iteration) the process of oternation continueing until all positions are fillied in a single iteration.

A reiterative count also allows for a reduction in the quota as a result of exhausted ballot papers. (Votes that do not record a preference for a continuing candidate) as the quota is recalculated. The method of segmentation seriously distorts the proportionality and results of the election.

If you recount the Queensland 2007 Senate vote excluding all candidates other than the last seven standing (3 ALP, 3 LNP and 1 Grn) then Larrisa Waters would have been elected. The reason she was not elected is due to the method of segmentation. It is outdated an not necessary. With the aid of a computer count a reiterative count is easily.

with the aid of a scomputerised countomng system there is no excuse for not implementing a reiterative counting system. It is simpler, more accurate and more refective of the voters intentions.

Antony Green thinks that the public is too stupid and cannot digest this information and as such he prevents this issue from being openly debated.

It is not the role in the ABC or antony Green to decide what should be discussed and what should be restricted information. I thought we lived in a free democratic society) The ABC is there to facilitate public debate and open discourse. there is not justofcation for the ABC to prevent the publication of this information.

Information should be freely available and the reader and public should then be able to make an informed decision.