A fair voting system

Preferential voting

If a candidate in an election does not achieve a majority of first preference votes, the winner is determined by the allocation of subsequent second, third and so on preferences.

How Preferential voting works: 

There are two systems of preferential voting: Full preferential and optional preferential voting.  With full preferential voting, voters are required to indicate their first preference by placing a “1” against a candidate’s name, then make a second preference and so on for the number of candidates on the ballot paper.  Optional preferential voting only requires the voter to make a first preference.

If a candidate does not get an absolute majority of first preference votes, then the candidate with the least number of votes is eliminated and those votes are allocated to the other candidates according to the number of second preference votes.  If no majority has been achieved, the next candidate with the least number of primary votes is eliminated and those votes are allocated to other candidates according to the second preference or third preference and so on if the second preferences have been exhausted.

Supporters of preferential voting say:

  1. The winning candidate is the most preferred or least disliked candidate by the entire electorate.
  2. Voters who support minor parties know that their votes will count towards deciding the winner.
  3. Parties sharing overlapping philosophies and policies can assist each other to win.

 Opponents of preferential voting say:

  1. Vote counting is complex under current manual procedures.
  2. The process is costly and time consuming, potentially delaying a result.
  3. Some people don’t like having to choose more than one candidate.
  4. Preferential voting makes voting more difficult.  Some people do not like having to rank their preference of candidates.  They either neglect to do so or make mistakes, leading to higher levels of informal voting.
  5. Some people do not like being forced to make a preference for candidates they do not support.
  6. A candidate not supported by most of the electorate could still win.

What do you think?  Have your say.  Join the conversation below.

 

Drawing of Electoral Boundaries

Sometimes people find themselves in a new electorate when voting comes around.  That’s because electoral boundaries - approximately 100,000 people within an area - are sometimes changed to reflect changes in the movement of people and the demographic makeup of the area.  Electoral authorities regularly hold hearings to review boundaries. Political parties are not allowed to participate in the hearings so as to avoid the perception of manipulation of the system in their favour.  Some people do not think electoral boundaries are being decided fairly.

What do you think?  Have your say.  Join the conversation below.

 

How we vote 

Postal voting

Postal voting is designed for people who cannot attend a polling place in their electorate.  

How voting works:

Once a person has voted, the ballot paper is placed in a sealed envelope which does not contain any voter identification, and then is placed in another sealed envelope that contains the name and address of the voter.  When it is received by the electoral authority the outside envelope is used to confirm the person has voted. The ballot paper is removed from the inside envelope and place in a pile for counting. The system is designed so the identity of the voter cannot be linked to the ballot paper, thus ensuring tick the person’s vote is anonymous.

Opponents of postal voting claim that the system is open to abuse because votes can be tampered with and there is nothing stopping the voter’s personal details being copied. 

What do you think?  Have your say.  Join the conversation below.

Early voting

Early voting is officially known as 'Pre-Poll' voting -- voting before the actual day of the election or poll.  When voting early, voters are required by law to give a valid reason for their request to vote before election day.

How early voting works:

For the 2019 federal election, early-voting or pre-poll voting centres opened in each electorate three weeks before election day in metropolitan areas and two weeks before election day in rural areas. 

According to AEC figures, 2980498 people voted early for the 2016 federal election.  In 2019, 4766853 people voted early -- a 60% increase in early voting compared with the 2016 election. 

Vote Australia recognises that early voting is convenient for those who need it.  Should all voters be allowed to vote before all issues have been fully debated?

What do you think?  Have your say.  Join the conversation below.

COMMENTS

Please check your e-mail for a link to activate your account.
  • Peter Newland
    commented 2021-08-23 20:54:27 +1000
    Troy Arnould: In answer to your question: it is secret voting. However, one of the safeguards against electoral fraud protocols I proposed could allow a person to choose to reveal their vote but only by choosing to have a receipt printed and then admitting that the receipt in their possession was their own vote receipt. But nothing on the receipt identifies the voter: it only identifies the place of voting and a unique random number that does not even identify the time of voting. That’s all now detailed in the url link spelled out in words in my previous comment, but I updated security recommendations some weeks ago.

    Also, since my original comment, I have better understood and explored the theory behind what began as an empirical approach that worked – but I wasn’t sure why. All I knew was that it was wrong to be obsessed with 1st preferences when sometimes last preferences were more important. Now I have changed the name from APR to DCAP. Candidate Acceptance Percentage is, in FACT, the exact Percentage of voters who Accepted the Candidate as being in the top 50% of Acceptable Candidates.

    Further, I was surprised to prove that Arrow’s Impossibility Theorem is wrong, and that my DCAP counting method guarantees to give a fair result in all cases. DCAP gives the same results as the Borda count of a FULL preferential voting election, but DCAP expresses the result in a consistent user-friendly way.

    In addition, DCAP easily and fairly handles PARTIAL Preferential Voting as well as SPLIT-Partial preferential voting. SPLIT Partial preferential voting makes it far easier for voters faced with, e.g., 123 candidates from 31 Parties to choose from to elect 6 Federal Senators for NSW. Currently, voting 1-6 for your top 6 Parties risks having zero say over who wins Senator 5 or 6 and hence who may hold the balance of power. But voting 1 to 31 Parties in order of preference is very difficult, let alone preferencing 123 Candidates. in order of preference. To overcome this, a SPLIT vote allows voters to vote (e.g.) 1-7 for their favourite Parties and 25 to 31 for their least favourite parties without having to preference the many micro parties they know nothing about. DCAP allows split votes to be counted fairly, and even to automatically correct voting errors where a voter’s intention is clear.

    DCAP is totally immune to strategic voting: even if a voter voted 1-7 and 93 to 99 for 31 parties, DCAP algorithms automatically translate 93 to 99, to 25 to 31. So, voters collectively get what they voted for, with no strategies available to distort a vote. However, no vote is ever immune from Party, Media or Government misinformation and propaganda.

    Borda is often criticised (by Arrow theory, and by numerical example) as being capable of ignoring an absolute majority. The same applies to DCAP and to its previous guises as: APR (Average Preference Rating); WPC (Weighted Preference Counting); or, CPV (Consensus Preference Voting). Until recently, I conceded that an absolute majority should override WPC/APR/Etc.
    I was wrong.
    Borda/WPC/APR/DCAP override very narrow absolute majorities ONLY, repeat O N L Y when that is in fact the BEST result. This can only happen when the ‘winner’ has a minuscule absolute majority margin compared while that ‘winner’ more strongly deserving the ‘wooden spoon’ as the MOST DISLIKED candidate on preferences.

    Other ‘counting’ methods suffer from erratic tipping points and that’s why they can and do get it wrong – often when it matters most in determining the last one or two candidates elected in multiple representative electorates where such winners often hold the balance of power. In contrast, Borda and DCAP are totally ‘linear’: i.e., they have no sudden tipping points, where changing one vote can tip preferences towards a totally different candidate. This is because Borda and DCAP never eliminate candidates, never distribute preferences, they fairly take into account all preferences for all candidates, in filling all vacancies and every vote has the same say over every vacancy.

    Current eliminate & distribute ‘counting’ methods are inherently flawed: they can not guarantee fair results; plus they facilitate strategic voting trying to game the system.
    Further, some claim that partial preferential voting (but not SPLIT partial preferential voting) skews the system in favour of larger parties. In contrast, DCAP actually guarantees a “fair voting system”.

    Why the D in DCAP?
    DCAP results can seem extremely counter-intuitive. E.g.: With 4 candidates, DCAP will correctly declare candidate D, with ZERO first preferences but 100% 2nd preferences, as the CLEAR STRONG DCAP winner with a DCAP score of 66.67% despite: Candidate A getting an ‘absolute majority’ with 51% of first preferences, but with a DCAP of only 51% due to A’s getting the largest share, 49%, of 4th-or-last preferences; and Candidate B getting 25% of both 1st and last preferences; and, C receiving 24% of first preferences.

    Although absolutely correct, this DCAP result is counter-intuitive and contrary to Arrow’s Impossibility Theorem. Hence the ‘D’ in DCAP to acknowledge that my D’Nalwen Certainty Theorem and DCAP guarantees a fair election counting method, and disproves Arrow’s so-called “Impossibility Theorem”, D’Nalwen being my surname in reverse. Full details are in the url link in my previous comment.
  • Troy Arnould
    commented 2021-08-23 09:50:39 +1000
    Does it allow for secret voting or are all votes associated to a user and searchable/traceable?
  • Cayden
    followed this page 2021-06-15 10:17:44 +1000
  • Peter Newland
    commented 2021-06-04 08:44:15 +1000
    The statement: “The winning candidate is the most preferred or least disliked candidate by the entire electorate” is simply incorrect as can be easily proved. The fact is that the preferential voting system is seriously flawed because it effectively ignores last and near-last preferences.

    Now preferential VOTING is the best system available. Period. However: the so-called ‘counting’ method is seriously biased towards 1st preferences while totally ignoring last preferences. The result is that a candidate can “win the wooden spoon” as being the MOST DISLIKED candidate by a healthy margin and YET be declared the ‘winner after preferences’ by a narrow margin.

    This is proved in simple numerical examples at https colon slash slash tinyurl dot com slash ElectoralReformOz all one word. This is not just an obscure theoretical problem: it is a real issue that can and does gets candidates elected AGAINST the voters preferences. The above link shows that 3 of the 25 MLAs elected in the 2020 ACT election were elected against voters preferences – and the link names those wrongly elected and those wrongly eliminated including giving the exact counts showing how the eliminated candidate won against the ‘elected’ candidate in a virtual one-on-one run-off election as determined by the actual preference votes at the ElectionsACT website.

    The link also introduces the authors Average Preference Rating (APR) system of fairly counting preferential votes to guarantees to always exactly follow whatever preferences voters have marked. APR allows SPLIT Partial Preferential voting which allows voters to mark their first few, and their last few, preferences: the SPLIT option makes it easier for voters faced with ballot papers such as vote 1-32 Above the Line, Or 1-140 below the line.

    APR is very VERY fast to count provided that all votes are SCANNED &/OR DIGITISED. Votes could be scanned on site and a VERY small file uploaded from each polling place and amalgamated very quickly at the main electoral office. The reason for the small file size and speed of counting is that no votes are ever distributed, every preference mark is taken into account and the system is totally fair. The system easily allows multiple parallel audit trails that make election fraud extremely difficult – it ist even possible to allow a voter to check that their particular vote was actually counted and recorded correctly in the count by using unique random vote IDs generated when the voter votes – plus candidates, parties and media could get near real-time info re individual polling booths or electorate tallies as votes are counted. The system allows even a close Australian Federal Senate Election results to be available within hours of polls closing.
  • Peter Newland
    followed this page 2021-06-04 08:44:11 +1000

JOIN THE FIGHT

Add your name and let’s keep Australian elections free and fair.

  • commented on A fair voting system 2021-08-23 20:54:27 +1000
    Troy Arnould: In answer to your question: it is secret voting. However, one of the safeguards against electoral fraud protocols I proposed could allow a person to choose to reveal their vote but only by choosing to have a receipt printed and then admitting that the receipt in their possession was their own vote receipt. But nothing on the receipt identifies the voter: it only identifies the place of voting and a unique random number that does not even identify the time of voting. That’s all now detailed in the url link spelled out in words in my previous comment, but I updated security recommendations some weeks ago.

    Also, since my original comment, I have better understood and explored the theory behind what began as an empirical approach that worked – but I wasn’t sure why. All I knew was that it was wrong to be obsessed with 1st preferences when sometimes last preferences were more important. Now I have changed the name from APR to DCAP. Candidate Acceptance Percentage is, in FACT, the exact Percentage of voters who Accepted the Candidate as being in the top 50% of Acceptable Candidates.

    Further, I was surprised to prove that Arrow’s Impossibility Theorem is wrong, and that my DCAP counting method guarantees to give a fair result in all cases. DCAP gives the same results as the Borda count of a FULL preferential voting election, but DCAP expresses the result in a consistent user-friendly way.

    In addition, DCAP easily and fairly handles PARTIAL Preferential Voting as well as SPLIT-Partial preferential voting. SPLIT Partial preferential voting makes it far easier for voters faced with, e.g., 123 candidates from 31 Parties to choose from to elect 6 Federal Senators for NSW. Currently, voting 1-6 for your top 6 Parties risks having zero say over who wins Senator 5 or 6 and hence who may hold the balance of power. But voting 1 to 31 Parties in order of preference is very difficult, let alone preferencing 123 Candidates. in order of preference. To overcome this, a SPLIT vote allows voters to vote (e.g.) 1-7 for their favourite Parties and 25 to 31 for their least favourite parties without having to preference the many micro parties they know nothing about. DCAP allows split votes to be counted fairly, and even to automatically correct voting errors where a voter’s intention is clear.

    DCAP is totally immune to strategic voting: even if a voter voted 1-7 and 93 to 99 for 31 parties, DCAP algorithms automatically translate 93 to 99, to 25 to 31. So, voters collectively get what they voted for, with no strategies available to distort a vote. However, no vote is ever immune from Party, Media or Government misinformation and propaganda.

    Borda is often criticised (by Arrow theory, and by numerical example) as being capable of ignoring an absolute majority. The same applies to DCAP and to its previous guises as: APR (Average Preference Rating); WPC (Weighted Preference Counting); or, CPV (Consensus Preference Voting). Until recently, I conceded that an absolute majority should override WPC/APR/Etc.
    I was wrong.
    Borda/WPC/APR/DCAP override very narrow absolute majorities ONLY, repeat O N L Y when that is in fact the BEST result. This can only happen when the ‘winner’ has a minuscule absolute majority margin compared while that ‘winner’ more strongly deserving the ‘wooden spoon’ as the MOST DISLIKED candidate on preferences.

    Other ‘counting’ methods suffer from erratic tipping points and that’s why they can and do get it wrong – often when it matters most in determining the last one or two candidates elected in multiple representative electorates where such winners often hold the balance of power. In contrast, Borda and DCAP are totally ‘linear’: i.e., they have no sudden tipping points, where changing one vote can tip preferences towards a totally different candidate. This is because Borda and DCAP never eliminate candidates, never distribute preferences, they fairly take into account all preferences for all candidates, in filling all vacancies and every vote has the same say over every vacancy.

    Current eliminate & distribute ‘counting’ methods are inherently flawed: they can not guarantee fair results; plus they facilitate strategic voting trying to game the system.
    Further, some claim that partial preferential voting (but not SPLIT partial preferential voting) skews the system in favour of larger parties. In contrast, DCAP actually guarantees a “fair voting system”.

    Why the D in DCAP?
    DCAP results can seem extremely counter-intuitive. E.g.: With 4 candidates, DCAP will correctly declare candidate D, with ZERO first preferences but 100% 2nd preferences, as the CLEAR STRONG DCAP winner with a DCAP score of 66.67% despite: Candidate A getting an ‘absolute majority’ with 51% of first preferences, but with a DCAP of only 51% due to A’s getting the largest share, 49%, of 4th-or-last preferences; and Candidate B getting 25% of both 1st and last preferences; and, C receiving 24% of first preferences.

    Although absolutely correct, this DCAP result is counter-intuitive and contrary to Arrow’s Impossibility Theorem. Hence the ‘D’ in DCAP to acknowledge that my D’Nalwen Certainty Theorem and DCAP guarantees a fair election counting method, and disproves Arrow’s so-called “Impossibility Theorem”, D’Nalwen being my surname in reverse. Full details are in the url link in my previous comment.

  • commented on A fair voting system 2021-08-23 09:50:39 +1000
    Does it allow for secret voting or are all votes associated to a user and searchable/traceable?

  • signed up on Subscribe 2021-08-12 18:40:33 +1000

  • followed A fair voting system 2021-06-15 10:17:44 +1000