On Herding Cats

And now, away from outreach and on to internal governance. In the interest of saving labor, I will start with an essay I wrote for LP News that never got published:

In the course of spreading the Libertarian message beyond the usual suspects, I have attended the meetings of many other groups, ranging from the hippie left to the pro-business right. In all these meetings one common factor caught my attention:

They behaved.

They actually had members content to go to a meeting to mainly listen and vote. Some of the meetings of the hippie left were particularly striking: people forming circles, holding hands, and feeding off the group energy as everyone nods in agreement with the leaders.

No so with Libertarian meetings! Whenever two or three Libertarians are gathered together, there are three or four opinions on any subject, and everyone has a 500 word thesis ready.

This is fun over a few beers. It can get tiresome during a committee meeting with a dozen people in attendance – which gives only five minutes per hour per speaker. It is downright painful at conventions where discussion on any motion is quickly diverted to a discussion of an off the cuff amendment, or worse, discussion over which amendment is being discussed.

Running a meeting of Libertarians is like herding cats. No matter how well crafted the motion, someone will offer an amendment. No matter how much work is put in by a committee, there is someone who can do a better job from the floor. Meanwhile, some get bored and start mumbling or walk out to attend a breakout session. Others complain about such behavior. Then battles over procedure ensue…

There is no point in complaining or asking Libertarians to behave. We won’t. We are Libertarians because we don’t like group decision making. We don’t go along with the herd. We don’t defer to committees. We are cats; it is time to deal with this fact.

Robert’s Rules of Order are adequate for running a student government, a local Rotary Club, or for ratifying decisions already made in the back room. They work poorly for determining the consensus of a group of strong-willed intellectuals who don’t like being told what to do.

After suffering through many LP meetings ranging from county executive committees, state and national conventions, to the LNC, I have taken to researching and experimenting (at the county level) with alternative procedures. Here are some of the most promising:

1. Parallel consideration of amendments. A committee spends months crafting a platform or bylaws change. Within a minute of its presentation, someone on the floor has an amendment. Most of the time for debate on the committee’s report gets used on discussing and voting on amendments instead of on the merits of the motion proper. Worse yet, since amendments are considered on a first come, first serve basis, the most quickly (thus poorly) crafted amendments get consideration. Those who think before amending often get shut out by time considerations.

Why not let all variations on a motion be considered at the same time? This way people can compare the relative merits of multiple options, including the original motion. This also gives people who think before raising their hands a chance to offer their options. Finally, time currently wasted on points of order and procedural votes could be used productively.

2. Parallel voting methods. Robert’s Rules call for voting on amendments before voting on the main motion. This is a very serial procedure which disallows full consideration of all the options. Furthermore, people are allowed to vote for amendments to a motion even if they intend to vote against the motion as amended. This is a terrible conflict of interest! It would be far better to let each variation on a motion stand on its own merits.

The Free State Project did parallel voting on all the states under consideration using Condorcet’s method. This required all voters to rank each state from most favorite to least favorite. Then, a computer ran a head-to-head vote between each pair of states. The state which won all of these votes (New Hampshire) became the state they pledged to move to.

While Condorcet’s method is the gold standard of voting systems, it is too cumbersome for live meetings. A much simpler method is approval voting. There, voters vote up or down on each version of a motion. If none of the versions reaches the required threshold (such as majority or 2/3, depending on the issue), then the motion fails. If more than one version reaches the threshold, the version with the highest approval wins. The beauty of this approach is that the version which is disliked the least wins, which is very good for party unity.

The downside is that voters cannot express degrees of approval. We can improve on approval voting by having elimination rounds if more than one option meets the required threshold.

3. More seconds. Even with these improved procedures, things could get cumbersome if too many options are put up for consideration. One way around this problem would be to require a greater fraction of the body to second a version before it is considered for voting. Since voting follows consideration of all versions, this could be a background process, with members co-signing an amended version as debate proceeds.

4. Creative recess. When an entire convention is in session, in theory only one person gets to speak at a time. This limits communication bandwidth considerably. One way to allow more debate on contentious issues would be to put a recess between presentation and voting. This would give people more time for crafting amendments, soliciting co-signers, and presenting written versions to the secretary.

This process could be facilitated by setting up stations for each motion where people could continue debate and craft amendments. This is the idea behind Open Space Technology (www.openspaceworld.org), a hybrid of democratic and market based decision making which has been embraced by “democracy activists” around the world. I have seen this format in practice by local leftists. Ironically, the far left is embracing a meeting format that is more market-based than that which we Libertarians use at our conventions. Perhaps we should fix this.

Since I wrote this article, a better parallel voting method has been brought to my attention: Range Voting. With Range Voting, the voters grade each option on an agreed upon scale. A scale of 1-5 works well in a committee setting. Finer gradations require electronic tallies. The grades are averaged and the option with the highest grade wins. (In parliamentary settings, minimum grades may be required for special votes. Otherwise, the winning option simply has to beat the Nay option.) This voting system is often used in judging situations such as: sporting events like diving and gymnastics, GPAs for valedictorian and marching band competitions.

There is some powerful theory that shows this to be a superior system. See www.rangevoting.org for the theory. I have seen it successfully used in action under a different name (consensus voting). A 1-5 scale was used for determining consensus for the LP’s Strategic Planning Team. This worked so well that I used it for ranking essays at www.ReformTheLP.org. It was only later that I was told that this was called Range Voting and that it was backed up with strong theory.

Previous | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | Next

Copyright 2007, Carl S. Milsted, Jr. All rights reserved.