User:Average/Fractional voting
Bi-Valent, Fractional Voting is a method for ensuring the greatest amount of voice of the voters is heard and that no one gets elected to a position of power if no one believes enough in them. That means, it does work even if there's only one candidate, because it won't vote in an unsupported candidate. It is "bi-valent", allowing negative votes. This solves the (USA 2000AD) Nader problem of lost votes helping the rejected Republican candidate. These two techniques make it a cheaper voting method because it eliminates the need for "primaries" to weed out all possible contenders on a given side.
- Each voter gets a number of tokens equal to the number of candidates on the ballot. They are not obligated to use them all, if they don't feel they have enough information.
- The voter can allocate these tokens as they wish FOR or AGAINST. They can put all the tokens for a single candidate, against, or distribute them amongst any number of them.
- The winner is decided by summing all the values given to each candidate by each voter at the level of the position. That means that you can't just count the numbers at the city level and report the winner to the next, but you should report the tally across the board so that the fullest information is given for the decision. Later technology may allow cities to provide rounded data that is more personal to the district.
- If you have positive results, the highest one wins. A negative result indicates that there is not enough confidence in any of the candidates, or the timing is wrong and you should not decide.
Fractional voting can be adapted to assist with governance in hackerspaces where you want to maximize value from people, rather than value of governance. For this, each person should up use all their votes, and vetted members (perhaps board members, for example) can use additional decimal digits which can tip the scales over lesser voters (due to presumably better knowledge and greater precision). This could be generalized if you're implementing a ranking system for members to decide how many digits of precision they get to employ.
Total number of votes counted should be published or communicated (w/total voters if you kept track), but individual results should not be published except for simple ranking (1st, 2nd, 3rd).
The number of tokens should be equal to the number of candidates in the initial period. After the public has gotten used to the new voting method, it is superior to use double the number for finer-grained representation.
This voting system solves >4 major problems with existing voting methods:
- the expense and trouble of holding primaries as mentioned (via allowing votes proportional to # candidates and grouping of votes)
- the Nader problem as mentioned above (via grouping & negative votes)
- the election of incompetence problem (via rejection of candidates who don't get a positive total)
- the feeling of powerlessness of the voter (via fractional and negative votes)
- for government: prevents lifetime appointments of judges by imcompetent executives, making a higher-quality Judicial Branch.
