The U.S. election system is rigged against third parties. With Plurality Voting most third party candidates suffer the Lesser of Two Evils Dilemma. I have found a couple of loopholes. (Read the book!) But there may be an easier solution: change the system.
That’s right: easier!
Changing the Constitution is hard, so I’m not suggesting that we can easily fix Presidential elections. But states and localities have control of elections further down the ticket, and history shows that states and localities are open alternative voting systems.
Furthermore, election reform is a transpartisan issue. Here, libertarians, progressives, greens, good government types, conspiracy theorists, mainstream media, etc. can work together.
In fact, there is momentum happening right now — for the wrong reform. Maine has enacted a Ranked Choice method (specifically Instant Runoff) for future elections.
I could tell you in great depth why I think Ranked Choice is the wrong reform, but I won’t (for now). Instead, I wish to show you. Over at my older quiz2d site, I have created poll for the 2020 Presidential primaries using multiple voting systems: Plurality, Approval, Range, Ranked Choice, and Range with Runoff. Give it a try yourself. Notice the complexity of Ranked Choice: both for filling in the ballot and for viewing the results.
In the future I will post deeper arguments. For now let reality do the talking. Take the poll and encourage others to do the same.
An Easy Plan of Action
The time for lobbying is not yet. To change the system we need more people to understand the problems with Plurality and to fully understand the alternatives. My online poll is a start.
A good next phase is to encourage people to try alternatives in smaller groups: social clubs, fraternities, sororities, church groups, school clubs, corporate boards, etc.
To some degree we already use Range Voting in these situations. A voice vote is a Range vote. Loudness is a measure of passion. Not just head count.
Perhaps the simplest reform would be to replace Robert’s Complicated and Easily Gameable Rules of “Order” with Approval Voting when people want to amend a motion. One way to kill or pervert a motion is to quickly offer an amendment. Debate can run longer on the amendment than the motion itself. This happens frequently at Libertarian Party conventions.
One particularly dirty trick is to offer an amendment to make the main motion worse in hopes of killing it. We have some bad laws on the books from times when this dirty trick backfired.
Approval Voting makes it possible to debate and vote on multiple versions of a motion in parallel. Just have a vote on all the versions, allowing people to vote Yea on as many versions as they approve of. The version with the most Yeas wins. (Or Nay wins if none of the versions get a majority.)
If this is not good enough, Range Voting would be the next step. This works well in a small group, like a committee even without electronic tallying. I first experienced Range Voting when I was on the Libertarian Party’s Strategic Planning Team. It worked well. This says a lot, since the LP leadership is nearly as good at gaming parliamentary rules as the U.S. Senate.
For a bigger group, electronic counting would be the next step. If I had the time and/or the money to pay some developers, I’d write a web/phone app that clubs could use for electronic voting at conventions and other largish meetings. Better yet would be a free software project so clubs could own their databases. (A shared service concentrates power. Mmmmmmmm, power…)
When enough people have seen the benefits of better voting systems in practice, taking it up to the government level will be fairly easy.
Great demonstration, Carl! Thank you for the time and effort it took to set this up.
Besides gaming the system, I think another reason people might give zeros on range voting is to avoid scoring candidates they don’t know enough about. (In a large field of candidates, it will be difficult for the broad electorate to have informed opinions on each.) In such cases it seems an abstain option might be better than giving the candidate a zero?
Hi James.
Yes, people might well give candidates zeroes because they are unknown. That is reasonable.
Others might give unknown candidates nonzero scores because they despise some of the known candidates so much that they prefer the unknown to the bad known.
Back in the early days of the Center for Election Science, there was a great deal of thought put into how to handle abstentions. How do you count an abstain? What should the quorum be? Can we have negative numbers so people can flag whom they actively dislike?
None of the answers were satisfactory. The CES has since backed away from Range Voting in favor of the simpler Approval Voting, and I ended my involvement years ago.
My solution for the dark horse problem? Let individual voters decide. For example, I would take a random nutcase over Bill deBlasio, so I would give deBlasio a zero while giving unknown candidates higher than zero scores. Others may prefer to give the unknown zeroes by default.
—
The problem is even worse for Ranked Choice unless you have the option to rank only a subset of the candidates — which could work, I suppose. But only allowing one to rank a subset would be a disaster! That takes you right back to the Lesser of Two Evils problem. Even worse, it gives an edge to radical third parties over moderate third parties. Look at what nearly happened in France.