[EM] Re: Voting Systems Study of the League of WomenVoters of Minnesota

[Abd ulRahman Lomax]

At 04:54 AM 6/8/2005, James Gilmour wrote:
>Abd ulRahman Lomax Sent: Wednesday, June 08, 2005 5:01 AM
> > Let me make sure I understand. If we had a face-to-face  meeting, and an
> > election was held by show of hands, which is not an uncommon  thing, I've
> > never seen a rule that prevents a person from voting for more  than one
> > candidate. And the winner is the person with the most hands shown.
> > Essentially, approval voting is *standard*.
>
>My experience of face-to-face show of hands voting is limited to meetings 
>in the UK, but I have never ever been present
>at such an election in which a voter had more than one vote, ie voter 
>expected (or allowed) to raise his or her hand
>once only to vote for only one candidate.  The standard assumption has 
>always been that the election was a
>first-past-the-post plurality contest and that the winner would be the 
>candidate with the most votes when each elector
>voted for only one candidate.

You may not have noticed. I'm quite sure that in some such elections that 
I've witnessed, some voters raised their hands more than once, but nobody 
was keeping track. I don't think I ever noticed it. This will happen if the 
voter, for example, changes his or her mind after casting the first vote. 
And I've never seen a formal rule preventing it. It is a near certainty 
that some people will sometimes vote more than once in such elections. It 
might be rare, but given the number of such elections and the nature of 
people, I'm sure it happens. And what is not prohibited is allowed, until 
and unless someone objects and the matter is considered. (And I've never 
heard the "vote only once" rule stated in an informal vote.) Given that the 
process does not record individual votes but only totals, most times the 
exercise of such a right would go unnoticed unless someone noticed the 
individual voter or unless someone noticed that the total number of votes 
exceeded the number of members in attendance. Given that some people almost 
always abstain, even this clue would be generally missing. The only way to 
notice it in a meeting would be to pay close attention to who is voting for 
whom. Which is considered a bit tacky. Only when a secret ballot is held 
would the question arise. And why should secret ballots have a special rule?

What happens in informal organizations if someone marks more candidates on 
a secret ballot than the apparent intended number? I've never seen the 
question arise, though it easily could. Again, I'm pretty sure that it 
happens from time to time, especially in multiwinner elections.

My guess is that the ballots are ordinarily counted. The counters may 
grumble, but they may also realize that it is relatively harmless to count 
both votes -- it works no inequity in a single-office election, and it is 
not clear to me that it does in a multiple-winner election, i.e., if 
someone marks too many candidates. But I haven't considered the matter deeply.

Or the counters may discard the ballot. But if they did that, I'd guess 
they'd report it. I've never seen such a report in small organizations. 
Again, in the absence of explicit rules to keep or discard, we might never 
know what decision a counter made. In a multiwinner election, where a 
violation is more likely to take place (the more winners the more likely an 
overvote), it might not even be noticed, since counters would be looking at 
many ballots and would likely not explicitly count the number of total 
votes on each ballot. Rather they would be busy accumulating the totals. It 
would be quite easy to overlook an overvote.

The "standard assumption" in an informal show-of-hands election is just 
that. An assumption, not a rule. Organizations variously have majority or 
plurality rules and handle elections with no majority winner by holding a 
runoff. Which in an election by show of hands is quite easy to do.

And what would actually happen if someone raised their hand twice -- which, 
again, I expect, is not very uncommon -- and someone noticed? Would they 
interrupt the election with an objection? In an informal organization 
following normal parliamentary procedure, the chair would rule, and the 
ruling would be subject to appeal to the assembly. It is highly unlikely 
that the chair, in such a case, would actually toss out the vote; rather 
the chair would have two likely options: allow the vote, which would be 
simple and the quickest resolution, or retake the entire election, 
considering the first one spoiled. There would be pressure not to do the 
latter, and, if the election were complete, it would only be done if the 
vote changed the outcome. Except that someone else might object, "But what 
if there were more than one such multiple vote, and only one was noticed? 
Surely it is unfair to accept such an election while not accepting one 
where the vote is critical, since we did not have any means of reliably 
detecting such overvotes, and we only know about this one because of the 
complaint of Mr. X, who is obviously biased, since it was he who lost 
because of the overvote."

Under normal rules, debate is not allowed on such motions; but what I 
speculated as the text of an objection is not a debate, it is an objection 
on which the chair would rule immediately, subject to appeal to the 
assembly, which would again vote without debate. My guess is that, however, 
the sequence of events would stimulate consideration of the question at 
another time.

> > The oddity is the practice of
> > discarding ballots which are multiply-marked, as if they were somehow
> > defective. Does anyone know the history of that practice?
>
>For most of last century, most UK single-winner public elections were FPTP 
>simple plurality and to ensure that that
>system worked as intended, most ballot papers carried an instruction like 
>"You may vote for one candidate only".

Yes. U.S. ballots say the same thing, and voters who violate the rule have 
their ballots discarded. What I'm pointing out is this is at variance with 
open balloting by show of hands, and I'm asking why. I have not yet 
researched the law, which may vary from state to state. Nor have I 
researched the previously mentioned court precedent, where a city ordinance 
allowing a form of IRV (as I recall) was invalidated because it allegedly 
allowed people to vote more than once. I intend to. But being creatively 
lazy, I'm asking here....

>   Any
>ballot paper marked with more than one "X" was automatically discarded as 
>a "spoilt paper".  This is still the rule in
>all our FPTP X-vote single-winner elections.

Yes. Same here. I think it may have thrown the election to Bush in 2000, 
because of the Florida effect, and in particular because of the famous 
butterfly ballot, which caused many voters to incorrect mark the ballot for 
a third-party candidate whom they did not support. A certain percentage of 
these, quite understandably, perhaps not having time to go back and get a 
replacement ballot or not knowing that this was their right, attempted to 
correct the problem by punching their desired candidate's position. At the 
very least this would remove the erroneous advantage given to the 3rd party 
candidate (I think it was Buchanan), and it would also have restored the 
voter preference for Gore over Bush, if not for the rule preventing it and 
spoiling the ballot. Since counting the ballot would be closer to 
fulfillment of the Florida law requiring the counting of a ballot where the 
intent of the voter can be discerned from it, it might be argued that the 
rule spoiling such ballots conflicts with the law regarding counting, and, 
if so, it might be vulnerable to legal challenge.

>   If three candidates are to be elected by multiple X-vote (first three past
>the post), the voter may mark one, two or three Xs, but if the ballot 
>paper has more than three Xs it is automatically
>discarded as spoilt.

Yes. Standard practice. And why? Somebody tell me!

(I can probably guess. It is so done because it violates an expectation. 
Yet it does not seem to violate the legitimate purpose of the expectation, 
it only formally violates it. It is common for formal rules in 
bureaucracies to be applied outside their intention, it is one of the 
reasons people get thoroughly frustrated by bureaucracies! "Sorry, your 
application was rejected because you failed to write out the full four 
digit year, as we require, and instead wrote, for your birthdate, 05-05-35. 
You'll have to come back to the office to correct it.")

If there is any record anywhere of the debate leading up to the passing of 
the spoilage rule, it would be interesting to see if it was even 
considered, the effect of *not* spoiling such ballots. What damage would it 
cause? And would such damage, if any, be more serious than the damage from 
spoiling the ballot?