• Welcome to the new Internet Infidels Discussion Board, formerly Talk Freethought.

Maine could have avoided this with approval voting.

B

Blahface

Member
Joined
Apr 1, 2008
Messages
269
Location
Illinois
Basic Beliefs
Atheist
OK, so last election there was a ballot initiative in Maine to implement instant runoff voting statewide. During the this time some people (me included) were worried that this would violate the Maine constitution as it requires the winner to be the one with the plurality of votes. So, even though it passed through a referendum, the Maine Supreme Court gave the opinion that it violates the constitution.

So, it is a bit in limbo right now. The Senate voted to repeal it and the House voted to partially implement it. Clicky.

This could have all been avoided if they had went for approval voting instead.
 
This problem with approval voting is its violation of the "later no harm" criterion. It is, however, spoiler-proof and otherwise cloneproof.
 Comparison of electoral systems

Later-no-harm means supporting another candidate won't hurt one's original candidate. Later-no-help is like that, but with no helping instead of no hurting.

Being cloneproof means that candidates with similar voter appeal will not cause spoiling, teaming, or crowding:
  • Spoiling: they hurt each other
  • Teaming: they help each other
  • Crowding: they affect other candidates
 
I'll go through the list in that Wikipedia article.
  • FPTP (Plurality) -- Clones: spoilers -- LNH: not relevant
  • Approval -- Clones: good -- LNH: no
  • Range -- Clones: good -- LNH: no
  • Borda -- Clones: teams -- LNH: no
  • Maj. judgment -- Clones: good -- LNH: no
  • Runoffs -- Clones: spoilers -- LNH: yes
  • IRV -- Clones: good -- LNH: yes
  • Schulze -- Clones: good -- LNH: no
  • Copeland -- Clones: teams, crowds -- LNH: no
  • Kemeny-Young -- Clones: spoilers -- LNH: no
  • Minimax -- Clones: spoilers -- LNH: no
  • Ranked pairs -- Clones: good -- LNH: no
All the Condorcet ones violate LNH. So IRV wins here, in both clone resistance and LNH.
 
I am a fan of IRV too.

Concorcet systems always favor the most centrist candidate regardless of how actually popular that candidate is. I don't think you should get extra points for sitting in the middle even though nobody actually wanted you in the first place.

My issue with simple Approval voting is the later no harm criterion just mentioned by lpetrich. The only way I can boost my most favorite candidate over those I find only tolerable is to only approve of her and her alone. This soon degenerates into FPTP if enough people start thinking this way.
 
Here are some nice recent videos on voting systems.

Voting Systems and the Condorcet Paradox | Infinite Series
Mathematician Kelsey Houston-Edwards shows how a set of preference-voting ballots can give very different results when counted in different ways. That set has lots of circular preferences, so that may not be surprising.

She counted up first-past-the-post, two-round runoff, sequential runoff (IRV), and Borda. She also evaluated the ballots' Condorcet matrix and she found that the ballots have a Condorcet winner. All five winners were different.

Source for the ballots: Feature Column from the AMS

Arrow's Impossibility Theorem | Infinite Series
For preference voting, it is impossible to have a system that always satisfies certain conditions on the outcome.

CGP Grey - YouTube also has several videos on voting systems, like:
Quick and Easy Voting for Normal People (approval voting)
Politics in the Animal Kingdom: Single Transferable Vote
The Alternative Vote Explained (IRV)
Mixed-Member Proportional Representation Explained
The Problems with First Past the Post Voting Explained
Gerrymandering Explained
The Shortest-Splitline Algorithm: a Gerrymandering Solution [Bonus Video]
Multiple Party Gerrymandering [Bonus Video]
 
I completely forgot about this thread.

lpetrich said:
Disappointing. :(

However, approval voting is not that great for competitive elections -- it tends to degenerate into first past the post because voting for another candidate means giving that candidate an equal weight of support.
Why Approval Voting is Unworkable in Contested Elections - FairVote
New Lessons from Problems with Approval Voting in Practice – FairVote
Click to expand...

Those are a few cherry picked examples and I don't think they accurate representations of how voters will vote. One problem is that there is a selection bias. They are independents who probably aren't too concerned about voting for the lesser of two evils. Most people vote to keep the scary guy out of office and will make compromise votes.

If you look at the the French 2007 Presidential poll and the Maine 2014 Gubernatorial poll, you don't really see really see bullet voting. In both of those polls, they appeared to elect the Condorcet winner which IRV failed to do in the French one. It appears that voters from both sides will compromise and support the independent or guy in the middle.

To the extent that it does devolve into plurality – it will devolve into a more honest plurality in which there is no penalty for supporting your favorite.

Truth be told though, I do think the later no harm criterion is a problem. It is not a very big problem, but it is still a problem. I think most of it can be overcome by a runoff from a non-partisan primary. People might be more honest if they think there will be a second chance to vote for their favorite if they also vote for lesser liked candidates.
 
zorq said:
I am a fan of IRV too.

Concorcet systems always favor the most centrist candidate regardless of how actually popular that candidate is. I don't think you should get extra points for sitting in the middle even though nobody actually wanted you in the first place.
Click to expand...

The only problem I have with a Condorcet method is that it might accidentally elect a candidate because people will bury the guy they hate most with guys they don't know. I don't know how it would work in regular elections. I think it would be the best option for private organizations or for the House to elect the Speaker of the House, or for a parliament to elect a prime minister. Condorcet methods can be really useful for when the electorate knows the candidates.

Any time a candidate can get more votes than another candidate in a head to head contest, that candidate should always win. Otherwise, you are screwing over the majority of the voters. This is the problem that Condorcet methods solve. Condorcet methods also elect candidates in the center of the voters – not the center of Washington or even the center of the all the candidates running. It is the real center.

My issue with simple Approval voting is the later no harm criterion just mentioned by lpetrich. The only way I can boost my most favorite candidate over those I find only tolerable is to only approve of her and her alone. This soon degenerates into FPTP if enough people start thinking this way.
Click to expand...

I think a bigger problem is that under IRV it is not even safe to vote for your favorite. In fact, under IRV, it actually isn't even safe to vote for your second favorite – it is just that voting for your second favorite can't hurt your first choice. It doesn't mean that voting for your second favorite can't help your least favorite get elected.

The thing I like about approval voting is that it makes it much easier to vote on issues even if you don't follow politics. If you only care about legalizing pot, you can vote for all the candidates endorsed by the Citizens for Legalizing Pot. If you also care about net neutrality, just vote for all the candidates mutually endorsed by Citizens for Net Neutrality and Citizens for Legalizing Pot.

The big advocacy groups can be really effective. They can deliver large amounts of votes to candidates who pay attention to their issues. In such a competitive environment, candidates would have to pay attention to all the important issues.
 
You must log in or register to reply here.
Back
Top Bottom