In close follow-up with my desire to see political parties abolished, we also need to rethink how voting is done.
In the United States, you can only vote for a single candidate for most positions (town councils are an exception).
You do not have the opportunity to say anything more than a binary yes|no to a given person for a given office.
You can vote for Bob for mayor. But not voting for Mary, Quentin, and Zoe doesn’t really say anything about what you think of them – just that you liked Bob the best.
And there is the problem. There is an explicit elimination of relative preference when voting: all you can do is vote “yes” for a candidate.
That is very different from voting “no” against a candidate.
What should happen instead is you should vote for your favorite candidates in order of preference, so Bob is number 1, Zoe number 2, Quentin number 3, and Mary number 4.
Then when I vote, and rank them Mary 1, Zoe 2, Quentin 3, and Bob 4, we can get a picture of the relative preference of any given candidate running for the office.
Do this across all voters in a given election, and assign the winner to the person with the lowest score (in the numbering shown above – flip the values to assign the winner to the person with the highest score).
Perhaps even look at the top 3 or 4 after gestalt ranking, then vote again to determine the winner (this would be ideal for a Primary-then-General Election method).
What research shows is that while you and I may wildly disagree on “best” and “worst”, we’ll probably be pretty close on who we think is “good enough”.
In the Bob-Mary-Quentin-Zoe example with two voters, Mary & Bob both got 5 points. Quentin received 6, but Zoe earned 4.
The two voters, therefore, think Zoe is “good enough”, even though they part ways on “best” and “worst” (Bob & Mary).
Combine such a ranking system with a fully-open Primary election (ie you go rank every candidate regardless of “party”), and we would see much more representative-of-the-citizenry candidates appear at final Election.