In States without state-level initiative/referendum, conduct shadow primaries among candidates for the statehouse, filtered to just those candidates who agree to enact a splitting-resistant voting system.
Details
Can you imagine an existing or new organization, of whatever sort of tax filing:
- Organizing to operate, with allies if possible, in every statehouse district (for each chamber) of every State that does not have initiative and referendum to change voting systems at the State level,
- Soliciting campaign material from candidates (independent or of whatever party) for the statehouse who endorse, in principle, the establishment of spoiler-resistant voting systems,
- Republishing the campaign materials in one place online and in one print publication,
- Inviting the voters to a shadow primary among the filtered candidates,
- Exercising security such that voters can only vote once,
- Conducting the shadow primary using STAR or a system with similar qualities, and
- Announcing the results and encouraging all voters to support in the official election, the winner of the shadow election?
Expenses (which could get heavy)
- Election security
- Election logistics
- Publicity
Appendix 0 — Some Voting Systems
Based on what I am aware of as of Sat Dec 11 14:35:13 UTC 2021, I think there are grounds to expect that each of the following voting systems would resist spoiler effects strongly in the single-winner context. I list these by increasing complexity. Approval is arguably simpler than the current system, First Past the Post.
- Approval (for large elections)
- Score
- STAR
- Condorcet-compliant systems (maybe based on Llull/Copeland with some tiebreaking procedures added on)?
The thinking about what systems work well against vote splitting could evolve over time. Efforts should keep that in mind and plan to update. Also, which systems are permissible may vary according to State constitutions.
Appendix 1 — Summability by the Precincts
The Condorcet systems, while perhaps suitable among people with enough trust in one another to rely on computers, might not be safe for use in official elections. These systems require evaluating how candidates do against each other in their pairs, and when there are N candidates, there are N(N-1)/2 pairs. So, for example, 20 candidates leads to 190 pairs. Counting them could be a lot of effort without computers. And the US public does not, and should not, trust computers in an official election. Extreme security and transparency are merited in order to convince Trump supporters that there is no steal. Extreme security and transparency can be achieved with Approval, with Score with no more than about seven values in the range, or with STAR.
A reason to include Condorcet systems in the list is to head off the advocates of what they call “Ranked-choice Voting”. One can ask them, OK, if you like ranking systems, why do you prefer that one system you are promoting [video] over other ranking systems?
1787regime 