Easy fixing Voting
[I forgot to post this post during the elections, but still relevant]
Context : Two round uninominal ballot
At this moment in France, it’s the election period. And between two existential anguishes about AGIs, I sometimes think about the presidential campaigns.
We use a two round uninominal ballot: in each round, each citizen votes for a single candidate.
Problem : Bad properties
The uninominal voting system really does suck because primaries are not mandatory, so some candidates who could ally themselves do not ally themselves. So you have to vote strategically, and that has a lot of bad properties.
A classic example that highlights the bad properties of the uninominal ballot is the case where the majority of the population is left-wing, but 10 left-wing candidates are running while only one right-wing candidate is running. It is then the right wing that wins the election because of the spoiler effect, i.e. the dispersal of the votes of the left wing candidates.
Another problem is that strategic voting prevents conviction voting. It is only interesting to vote for those who have a chance to win.
Easy fix : the vote by approval.
There are a lot of possible voting systems (ranking, randomized condorcet, bordas point ballot, majority judgment ballot...). Which one to choose? Condorcet voting is relatively popular on this platform, but the main problem is its difficulty to implement and its complexity to vulgarize. The voting systems with the best properties are also unfortunately difficult to use.
The French YouTuber Mr Phi proposes a more pragmatic goal: By changing the minimum of rules possible, how to improve the elections as much as possible? What would be a credible and feasible alternative in the near future?
Pourquoi notre système de vote est nul (et le moyen le plus simple de l’améliorer)[Why our voting system is bad]
The answer presented is very simple: instead of voting for a single person, you can endorse one or several candidates. In practice, it would be enough to remove the constraint of a single name in the envelope, giving the possibility to vote for several candidates at the same time in the envelope.
This solution is straightforward because we would keep all the rest of the election organization: we count the names, and who has the most wins. We could even take the opportunity to eliminate the second round.
Link with effective altruism?
In theory, approval voting would make it possible to vote for several parties at the same time, and thus to promote more ideas at each election. This seems particularly important for animalist or transhumanist parties, whose votes are systematically cannibalized by the larger ecological parties. Some left-wing parties would have obtained funding much more easily, thanks to the guaranteed reimbursement starting from the 5% threshold.
It would also help to fight against hyper-polarization: candidates would no longer have to criticize other candidates, but could concentrate on being the best candidates, thus creating a more positive atmosphere. This would be a partial solution to the “politics is the mind killer”.
Politics
An association keeps a list of applications of usage of approval voting, but this list is rather short. In the short term in France, Emmanuel Macron and the majority have no interest in focusing attention on approval voting. But what is surprising is that the smaller parties don’t talk about it that often either.
A petition currently exists in France to pass the Condorcet ballot, but this petition has only 200 signatories when 500k are needed for the petition to be considered.
Réforme de l’élection présidentielle par un scrutin mathématiquement juste—Plateforme des pétitions de l’Assemblée nationale[Reforming the presidential election with a mathematically fair ballot—National Assembly Petitions Platform]
The title of the petition is “Reforming the presidential election with a mathematically fair voting system”. The fact that the Condorcet voting system is so complicated to explain does not help. So I have created a new petition in favor of the approval voting system.
I encourage you to do the same in your respective countries.
The Center for Election Science is not just a random organization. Jameson Quinn (who wrote the most about voting systems on LessWrong) used to be a board member and OpenPhil funded them.
Aim to explain, not persuade: the arguments against Approval Voting should at least be mentioned (eg https://fairvote.org/new_lessons_from_problems_with_approval_voting_in_practice/)
Depressing turnout of other major candidates doesn’t go down in value, so I see no support for the assertion that the campaigns would be less negative.