The Case for Democracy in CADE

David A. Plaisted

July 21, 1996

I would like to begin by thanking the program chairs and the local arrangements chair for an excellent (so far) CADE conference. I would also like to thank those who contributed to the CADE democracy proposition, including those who endorsed it at the beginning, those who supported it later, and those who responded to our survey. I would also like to thank the AAR Newsletter for publishing articles about this issue and to thank the trustees for a chance to present this motion here today. We have encountered a significant amount of support for this measure, something over 40 known supporters, and little opposition so far. This specific proposal has been endorsed by nine individuals including myself. In fact, in my conversations with the trustees, I have gotten the impression that they are not completely opposed to the democratic proposal either. They have even made a democratic proposal of their own. And thanks to Alan Bundy for arranging to have this meeting moved to Wednesday.

It is appropriate to start by considering the significance of the CADE conference. This conference, being at the center of the field of automated deduction, has a considerable effect on the lives and careers of each one of us working in this area, as well as on the development of the field of automated deduction. The trustees have general charge over the organization and running of this conference, and have direct and indirect influence on many of its aspects. So the choice of how CADE operates should be of immediate concern to each one of us.

Perhaps a historical comment would give some perspective on this issue. In the previous system for CADE, in effect through CADE 11, the program chairs were chosen by the program committee, and each program chair had a term lasting for about two years. CADE has been fairly successful under this system. However, some of the program chairs felt that it was necessary to incorporate CADE in order to handle finances better. Therefore, the current system, which went into effect between CADE 11 and CADE 12, was adopted.

The current system of organization for CADE is significantly different from the previous system and fairly recent (about 2 or 3 years old). The current system has some good points. However, it also has a significant drawback. It is less democratic than the previous system, in that CADE is run by a group of about 9 or 10 trustees, composed mostly of past and future program chairs, and future program chairs are chosen by the trustees. The effect of this is that in the current system, all trustees are chosen by current trustees. The members and the program committee have no direct voice in these choices.

The current system has no democratic element except that the members can vote to change the bylaws. However, this is too difficult a mechanism for the members to express themselves, as I know from experience. It takes a lot of work to change the bylaws, both in order to mobilize support and in order to write the bylaws.

My interactions with the trustees have convinced me that the current system differs significantly from the previous one in other ways, as well. The group of trustees has an identity that lasts a number of years. One generally communicates with the trustees through the president, whose term lasts several years at least. None of this was true before. This gives the trustees a lot more identity and authority than in the previous system. When one deals with the current trustees, one deals with a body that is going to remain substantially the same for several years, at least. I believe that this results in more reluctance on the part of some people to disagree with the trustees.

I must say that I was quite surprised that the current system was adopted, because it takes all direct decision making power in the choice of trustees out of the hands of the membership and program committee, except for the ability to change the bylaws.

I would like to make a proposal to introduce a democratic system in CADE. This proposal has been developed in discussions with many people, including the trustees. It is similar to the systems of ALP, RTA, and ISSAC, all of which have short terms for the trustees (or their equivalent) and 2 or 3 elected representatives each year on the average. These systems are all running smoothly.

I am very sorry that this issue has taken so long to resolve. I expected that this issue would be presented to the trustees, accepted, and soon implemented, as it was in RTA. I never expected to have such a long and drawn out discussion over it.

In the proposed system which we will vote on today, the trustees will be elected democratically. For the sake of simplicity in describing this system, I will assume that CADE will meet once a year. In the proposed system, the trustees would have three year terms, three elected each year, making nine trustees in all. The trustees could also appoint other officers, such as a secretary and a treasurer, but these officers would not automatically be trustees. As in the current system, the trustees would choose the CADE program chairs and local arrangements chairs, but these chairs would not become trustees unless they were elected. The constituency for the election of trustees would be the members of the Association for Automated Reasoning. We would prefer to let all CADE attendees vote, but decided not to change this in order to avoid complicating the issue. The trustee election would be by a simple two-round system, in which the second round can be omitted by a vote of the membership. The proposed system also permits special elections, in which all members of the AAR can vote. These special elections would occur between CADE conferences, and would help to reduce the effects of locality caused by the possibly small percentage of AAR members at each CADE conference. Motions passed at business meetings would be binding on the trustees, but could be reversed in a special election, which can reach the entire constituency. Our general philosophy has been to change the current system as little as possible and retain its good features, while making it fully democratic. We have made every effort to accommodate the concerns of the trustees in the development of this proposal.

Organizations such as the Association for Logic Programming have a democratic system similar to the proposed one, and it has been running smoothly for years. It should be clear from this and many other similar organizations that a democratic system is simple and workable for CADE and would not lead to significant problems of any type. It should also be clear that the mechanism of conducting elections is not a significant problem, either. In fact, no valid objections have been raised against my proposal.

It is probably well at this point to discuss the reasons for the change to a democratic system. I believe that in the long run, a democratic system is best and safest and gives the easiest way to correct problems that can arise. I admit that a democratic system can be inconvenient at times, but history has often shown that it is well worth the trouble. However, these proposed bylaws will impose very little inconvenience of any kind on the trustees, but will guarantee a completely democratic system. Every detail has been designed to give the membership the maximum voice in the running of CADE.

In the current system, there is no regular procedure for evaluating program chairs. Whether good or bad, program chairs become trustees, and help to choose future program chairs. There is a mechanism for removing program chairs, but it would probably be used only very rarely due to the embarrasment it would cause, and even then, the current trustees would do the evaluation.

If we did get a bad group of trustees, then the only recourse the members would have would be to change the bylaws, a most difficult and time consuming procedure even if the trustees do not obstruct it. In a democratic system, the members can express their wishes in an election.

The current system also has the danger of getting a group of trustees that people are not happy with, but who by their influence are able to prevent the bylaws from being changed.

When the trustees have authority but are not elected, I think that the temptation to favor one's friends and one's viewpoints can be too great. I'm sure that we would all like to think that problems will never arise, but we need to be realistic and to realize that human nature is at times self-seeking. A democratic system helps to put checks and balances on this tendency. In addition, disagreements and disputes do arise, and a systematic mechanism for solving them is useful. Furthermore, in a non-democratic system, there is some pressure on non-trustees to go along with the trustees in order to gain their favor. Thus one can end up with politics, even without elections. This may not happen with a particular group of trustees, but the potential for trouble is there. Elections help to provide more checks and balances.

There are in addition other problems with the current system. For example, it is difficult even for a well-meaning group of trustees to know what the members want without elections, since people who are unhappy are less likely to speak up.

There are some specific events in the history of the current system that already indicate the potential for a divergence of interest between the trustees and the membership of CADE. These are instances where it is at least reasonable that the membership would have acted differently than the trustees if they had had a voice. I will list them as follows:

1. The current system took considerable effort and expense to set up, and the general membership did not have much knowledge of what was happening. The membership may have wanted to be better informed about this.

2. The system that was set up was nondemocratic. I expect that most people would have preferred a democratic system.

3. After all this effort, the trustees may have been reluctant to change the system. It is conceivable that the membership would have favored a more rapid change.

The best way to know if people are happy with the trustees and with the system is to have a secret ballot as a regular part of the system. Even if the trustee candidate wins, the size of the negative vote can be a useful indicator of mood.

There are some additional advantages to electing the trustees. A democratic system will have an immediate psychological effect, making it clear that the trustees derive their authority from the membership. I believe that this will be a beneficial influence.

In general, elected trustees will tend to be more responsive to the wishes of the membership.

Elections would also give a good group of trustees reassurance that the community was behind them. This would be an encouragement to them.

It is also important to realize that this issue is important not only for us, but also for its effect on other conferences. Other conferences have non-democratic systems, but their members might also vote to make them democratic if they were given a chance. In fact, I would recommend that all conferences should be made democratic.

The current trustees seem willing to let the democratic proposal pass. In fact, they have made a democratic proposal of their own, although it has some problems. We should seize the opportunity to change the bylaws while we have a chance, and not risk getting a group of trustees who could put many more obstacles in the way of such a change.

Let me briefly mention some of the problems with the trustees' democratic proposal. It has somewhere between 3 and 6 non-elected trustees as well as six elected trustees, reducing the democratic portion. It permits the trustees to cancel motions passed at business meetings and special elections, if they choose. This is especially disturbing for special elections, which reach the whole constituency via e mail. It has too high a proportion (66 percent) for changing the bylaws. The means proposed for conducting elections is a lot of work and not necessarily any better than my two-round system. Also, with two trustees elected each year, it is likely that the trustee positions would be filled with program chairs serving their first and second terms, reducing the chance for the membership to elect an additional trustee.

I would like to pay particular attention to the trustees' insistence that upcoming program chairs should be voting trustees even without being elected. This serves no real purpose, and diminishes the democratic aspect of their system. In my plan, the upcoming program chairs could participate in all relevant trustee discussions and have an influence even without a vote. I believe that many program chairs would be willing to give up their vote among the trustees in order to insure a fully democratic system. Furthermore, program chairs and the secretary/treasurer are voting members of the conference committee that has charge of their CADE conference. This is where the real decision making takes place for the planning of the conference. The loss of their vote among the trustees is a small price to pay to achieve a fully democratic system. Finally, there does not seem to be any reason why upcoming program chairs should have a vote in the choice of future program chairs. So we see that my plan will not inconvenience the trustees in any significant way, and there is no reason not to adopt it.

I believe that the fully democratic proposal is in every way the better of the two democratic proposals. I urge you and appeal to you to pass this democratic proposal. If a democratic measure does not pass, it will mean that the membership has to some degree lost a voice in the running of CADE. The trustees' proposal will only give the membership a partial voice in CADE.

The issue is really who is in control: the membership or the trustees. I believe that the membership should be in control. Let's heed the lessons of thousands of years of human history that has often shown the superiority of democratic systems of government. A democratic system doesn't mean 6 elected and 3 or more non-elected positions; it means that all of the positions are elected.

For me, whether this measure passes or not is not the main thing, though I think it has a good chance to pass. For me the most important thing is to be on the right side, which I believe this is. Whether this passes or fails, I won't be ashamed to have fought to make CADE fully democratic. I encourage you to join me in this effort.

I believe we are at a turning point in the history of CADE. The current trustees will pass off the scene, but the organizational structure of CADE will last a long time. If we do not make the change to a democratic system now, the current system will become more and more entrenched and harder to change. It's unlikely that anyone else will go to the considerable effort required to change to a democratic system. If we wait until a serious problem develops, then we will have to deal with the problem and also try to change to a democratic system at the same time, which could be very difficult to do. Now when things are running relatively smoothly is the best time to make the change. If we do change to a democratic system, then it becomes easy to make further changes, since the trustees will have an additional reason to be responsive to the membership. So the proposed system can be regarded as just a first step, and if improvements need to be made, corrections can be made later. The issue is not whether you think the trustees are doing a good job - if you think that, you are free to vote for them under a democratic system. The issue is not the personnel, but the system. I encourage you to vote in favor of the fully democratic bylaws.