Problems with vote-discarding thresholds, esp in MMP
When thresholds exist for the purpose of limiting small parties, and when they work by discarding votes for small parties, they create perverse incentives for voters and parties alike.
Voters do not like voting for small parties because their vote is wasted and plays no part in decidig the makeup of the assembly. And often in MMP, in NZ esecially, the balance between left and right can be close. A vote for a small party is a vote lost to that party's allied parties. So small parties have no friends.
So the solution is to give the voter a second choice of party vote.
By doing that every voter can express support for a smallnew party and still have a say in the overall result. More details at ott.nz.
Keith Edmonds last edited by
@frenzed This is the classic case of vote splitting you get when you have a choose-one ballot. Your idea to give a second option would reduce this but not totally since their two favourites could both be small parties.
This is only one side of the vote splitting effect. There are several others that I do not need to cover here. As we know even moving to a full ranking would not remove all such effects since the centre squeeze effect would still be there.
To totally solve this one would need to move to a cardinal/score system. This is the revolution which happened in the last two decades in multimember systems. Multimember cardinal systems have basically replaced STV as the default system.
I see no reason why nobody has bothered to design an MMP system where you use scoring rather than choose-one. It could function just the same as the existing MMP style systems. You first run the election of the candidates. Then using those candidate's parties you compute the under served parties and award seats sequentially. This the formula for this could be modelled after RRV or SSS. There may be a nuance I am missing here but I am sure if somebody put in the time a full new MMP system could be made.
@frenzed You up for building such a system?
@keith-edmonds Thank you for your reply.
The idea is not that the voter identify their two favourite parties.
The question for the second choice is:
If your first choice party does not pass the threshold, to which other party should your vote go? (Just as with one choice, if for your second choice you choose a party that does not meet the threshold, your vote does not count in allocating seats)
This is a second choice offered in order that:
- A voter for a small party may also participate in seat allocation, irrespective of whether his preferred party passes the threshold.
- The true level of support for small parties may be known, and not influenced by strategic voting.
- Small parties do not act as spoliers. With one choice, small parties take votes from allied parties, and then see those votes go to waste.
- There are reduced barriers to the entry of new parties.
- The result is very close to proportional amongst votes for qualifying parties.
It is designed so that the voting paper requires minimal change, and vote counting is minimally more complex.
When measuring party support, mutliple ranked choices are not needed. Choose one (of the threshold passing parties) does the job.
This change acknowledges that parties are an essential part of national politics, allows smaller parties, and reduces barriers to entry. Parties need to present widely acceptable groups of candidates. If they don't, other parties can rise to take their place. And barriers to entry have been lowered.
MMP directly measures party support, so I do not see much value in scoring parties. MMP has the issue of overhangs not yet well-solved, but otherwise is well-liked in NZ. There is no mood to change from MMP.
Keith Edmonds last edited by
@frenzed I understand the system you propose and the justification for it. It does however have the flaws I brought up.
I do not think moving to a score ballot instead of a "rank two" as you propose is a much bigger change. It would give better results and completely solve the issue you are concerned about.
That's the same idea the German NGO Mehr Demokratie proposes. I proposed a similar modification to MMP that uses approval voting for the direct mandate and multi party choice. The party vote is distributed equally among all approved parties, then all parties that don't reach the threshold are excluded and the vote is distributed equally among the remaining parties.
It has the same effect as a second choice, but with the added benefit that voters don't have to limit their vote to one party.