Slate vs. individual voting, and how it pertains to fencing

[PeterGustafsson]

Hi!


Yet another thread on voting from me. You have been warned.

In the thread about suggestions from an USFA member on how to modify the USFA charter, he suggests that slate voting should be kept from happening, and that candidates should be considered on their own merits.

That is a legitimate concern, it is perfectly possible that a weak candidate hangs on to the coattail of a otherwise strong slate.

OTOH, with totally individual voting, there is always the risk that the elected group of officials will become unbalanced in one way or another. One region may be heavily over/underrepresented. The same goes for weapon/gender/age, other characteristics which it may be good to have some balance of. Here, I use the word "balance" as meaning that the proportion of those in the the elected group should not be too far away from the comparable proportion among the voting members. With a slate voting system, this can be counteracted.

I propose for your discussion a system which is intended to combine the best parts of both systems, while mostly avoiding the drawbacks. Consider this method:

1. For each charateristic (region/weapon/athlete status/gender/etc) minimum, and in some cases maximum, quotas are set. The quota is defined as a percentage of the elected group which has some charracteristic. The span between minimum and maximum permissible quota should be sufficiently large so that steps #2-9 can be carried out.
2. All who wish to run for election must state so, publicly, to the body in charge of electoral count. This should be done well in advance of the election, at least 6 months. The contenders state which posts they can consider running for, but they are not permitted to state whom they want to run with. They must also state which quota-specified characteristics they possess.
3. The electoral body publishes all information about the candidates obtained in step 3. This should be done at the very latest one month after the candidacy deadline, so at least 5 months before election.
4. All interested members are allowed to submit slates to the electoral body. Members are allowed to submit multiple slates, if the so wish.
5. The slates which fulfill all maximum and minimum quotas are suggested for popular election by the electoral body.
6. If a candidate finds that he is put up on a slate together with other candidates that he does not wish to be elected with, his only recourse is to step down from all candidacy. If that happens, all slates containing his name are stricken from the election list.
7. All slates surviving step 6 are put up for popular election. The names for the slate submitter is not divulged by the electoral body.
8. Members vote by stating their preferences for the different slates, 1 denoting the slate that they like the best, 2 denoting the 2nd best, and so on.
9. Should there be no slate which gets more than 50% of all the votes for best slate, the electoral body will proceed to find the best candidate by striking from contention the slate which got the fewest #1 votes. The votes for that slate are then transferred to whichever slate that the individual voter stated was his 2nd best choice. The votes are then recounted to find if any slate passes the 50% limit. Should this not produce a winning slate, the procedure is repeated sufficient times.

This voting method, Instant Runoff Voting (IRV) can be proved to always produce a winner. It is used by the IOC to determine which city shall host the games. Provided that the voting group is numerous and not intimately known on an individual basis by the candidate, it can be shown that IRV punishes negative campaigning, and rewards reach-out initiatives. It often produces relatively centrist election winners. There are no wasted votes, you can vote for your best choice as #1 and if it fails, you can have the the best to your liking of the popular choices as #2.

With this proposed system, grossly unbalanced elected groups would not win. People who only want yes-sayers around them will likely fall off at step #6. Widely popular candidates, and those who are very good for a specific post, will appear on many slates, and will therefore have a very good chance of getting voting. The system allows for negative campaigning, but it is unlikely to be of any use, and is very likely to be counterproductive. "Stealth candidates" will not pass step #2.

Apart from being more work-intensive than the present system (of an electoral committee putting up a slate which is accepted as a whole, unless an election is specifically called for) do you see any drawbacks compared to the present system? Any other comments?

I find it interesting that the suggestion from the chicago guy includes a big quota on people with a specific education, but no safeguards against inbalance in regard to other characteristics.


Have a nice time!

Peter Gustafsson

[davesaint]

Peter,

You need to adjust the proposal from the Chicago Guy. Individual Candidates, with a Vote Early, Vote Often structure. And remember in Chicago, even dead people can vote.

[oiuyt]

Note that the "Chicago guy" is from Columbus, Ohio. Wrong city, wrong state.

I believe that one of the intentions of the current system is to have actual elections -- which are a time- and expense-consuming process -- be extremely infrequent. Your system forces an election -- and a particularly involved one -- every election cycle. While that's clearly best if we're actually intending to use elections as elections, I don't believe that goal is one that is currently intended.

Note that IRV (which, in general I like) does NOT always result in the most-preferred result winning. Assume that there's a slate that is EVERYONE's #2 choice. It's the first slate eliminated, even if it occurs ahead of every other possible slate on more than half of the ballots (which it does assuming no slate wins in the first round). Regardless of what slate eventually wins, there's a slate which a majority of the voters preferred to the winner.

Another downside, assuming that there are multiple candidates for several positions, is that there could very easily be a huge number of barely-distinguishable slates. As a voter I might not care at all which of several dozen slates gets chosen (there's a few positions where there are several acceptable candidates without a strong preference), but I want any of them more than any of THAT several dozen slates, and THOSE several dozen slates I very much dislike. I now have to rank potentially hundreds of slates. This is a royal pain, can be confusing, and is prone to omission, errors, and other similar issues.

Even being able to easily sort the slates can be a confusing and difficult task in this situation. Especially depending on the media in which the slates are presented (trying to do this with only a physical medium can become prohibitively difficult).

-B

[PeterGustafsson]

Hi!

Originally Posted by oiuyt

Note that the "Chicago guy" is from Columbus, Ohio. Wrong city, wrong state.

Oops! Goes to show that one should check facts and proofread. At least it was the midwest! <-PG

Originally Posted by oiuyt

I believe that one of the intentions of the current system is to have actual elections -- which are a time- and expense-consuming process -- be extremely infrequent. Your system forces an election -- and a particularly involved one -- every election cycle. While that's clearly best if we're actually intending to use elections as elections, I don't believe that goal is one that is currently intended.

This could be alleviated by having the election done not by the general membership, but by a directly elected representative body - parliamentalism, if you will.

In the Swedish Fencing Federation, the voters are these:
1. Each dues-paying club sends voting representatives. If the membership<100, they have one vote. If the membership is 100 to 199, they get 2 votes. If the membership is at least 200, they get 3 votes.
2. Each of the 5 regions gets 2 votes.
3. The elected leadership gets 1 vote per member. They do not vote for the incoming leadership, but in all other issues put up for vote at the yearly federation meeting.

Proxy votes are allowed, but no representative to the yearly federation meeting may have more than 2 votes in total. (At one time, I voted for one slate of election committee with my own club´s vote, and for the other slate with the proxy vote of another south swedish club.)

The great majority of the time, those suggested by the election committee get elected, but they are always put of for election.

Since a national federation meeting is needed anyways, this election system does not introduce any new costs. Would it be viable in USFA? Would USOC allow it?

Originally Posted by oiuyt

Note that IRV (which, in general I like) does NOT always result in the most-preferred result winning. Assume that there's a slate that is EVERYONE's #2 choice. It's the first slate eliminated, even if it occurs ahead of every other possible slate on more than half of the ballots (which it does assuming no slate wins in the first round). Regardless of what slate eventually wins, there's a slate which a majority of the voters preferred to the winner.

I know that. You are in effect arguing for a Condorcet voting system. I also think that Condorcet is a good system, but it somewhat more difficult to explain to the mathematically challenged people (i.e. dumbasses) who have difficulties grasping matrix mathematics. Therefore, I went for something easier. In any case, both Condorcet and IRV are considerably better than first-past-the-post.

Originally Posted by oiuyt

Another downside, assuming that there are multiple candidates for several positions, is that there could very easily be a huge number of barely-distinguishable slates. As a voter I might not care at all which of several dozen slates gets chosen (there's a few positions where there are several acceptable candidates without a strong preference), but I want any of them more than any of THAT several dozen slates, and THOSE several dozen slates I very much dislike. I now have to rank potentially hundreds of slates. This is a royal pain, can be confusing, and is prone to omission, errors, and other similar issues.

Simple solution:
1. Find the 5 best slates according to your wish and rank them at the top accordingly.
2. Find the "dozen-or-so" worst slates according to your wish and rank them at the bottom accordingly.
3. Put in all other slates to places #6 -> #N-12 on your ballot according to list number, or lots, or whatever catches your fancy.

Originally Posted by oiuyt

Even being able to easily sort the slates can be a confusing and difficult task in this situation. Especially depending on the media in which the slates are presented (trying to do this with only a physical medium can become prohibitively difficult).
-B

If you really think that there will be lots of slates put up for election, one could limit the number by making the slate suggestor having to pay some reasonable sum (10$ ???) which must be sent in to the electoral committee before election. Make the sum high enough to deter those who act on a lark, but not a real hindrance to the really interested.


Have a nice time!

Peter Gustafsson