At 02:50 PM 4/26/2010, Peter Zbornik wrote:
Hi Jameson,
answers in the text.
On Mon, Apr 26, 2010 at 6:58 PM, Jameson Quinn
<<mailto:[email protected]>[email protected]> wrote:
Two questions, before I respond more fully:
1.
2010/4/25 Peter Zbornik <<mailto:[email protected]>[email protected]>
(v) asset voting is excluded due to lack of political support
Can you clarify? Is the problem with vote secrecy of the "lower"
delegates, and/or with the "back room" process among the "higher"
delegates (that is, the candidates in the current system)?
Yes that is indeed the problem - it allows for bribery and blackmailing.
You know, that's claimed, often. I've not seen a shred of evidence,
but it would depend on context. Bribery is something that can appear
when there is concentrated unsupervised power, and you are going to
get that if you have elections for officers who serve fixed terms.
It's pretty unlikely in your context. Blackmail is likewise, or
probably you mean vote coercion. Who would be coerced, the secret
ballot voters (Asset has a secret ballot initial stage), or the
electors, the candidates who hold votes. And why would not any kind
of delegate then be subject to these risks?
Tell me, in a political party the size you mention, someone tries to
coerce you into voting for them. What would you do? And what would
make you think that others would do differently? It would be
political suicide. (There can be literally crazy people who will do
anything, but that's a risk under all circumstances.)
The secret ballot was introduced together with the equal voting
right in many states of Europe, including the Czech Republic.
<http://en.wikipedia.org/wiki/Secret_ballot>http://en.wikipedia.org/wiki/Secret_ballot.
Notice that the protections of secret ballot are not absolute. Those
who count the votes can modify them, often, computers can be hacked,
etc. But under reasonably settled conditions, people can be and are
very open about their politics. I'm very skeptical that there is a
risk from Asset Voting as to bribery or coercion, that wouldn't exist
with any system you set up, practically.
There are versions of asset voting which avoid either or both
problem - the former, by only allowing votes for "qualified
candidates" (however that's defined), and the latter, by having
each candidate pre-declare their transfer order, which is then made
public simultaneously before the vote and used to automate the
transfer process. In other words, it's basically STV with one
predeclared ballot type per candidate.
The latter system is acceptable to me provided you can chose to cast
either an asset-type vote or a STV vote (in any case you can always
vote for yourself).
The latter system means that the preference orderings should be
clearly stated, which actually could be a good thing to make the
voting more transparent, but I wouldn't call it an essential part of the STV.
Asset Voting is STV, but with a chosen human being transferring votes
instead of a mechanical process set up by the voter. The problem with
the latter is that most voters don't have enough information about
candidates to do it well. The "declared" ballot variation is similar
to STV, and in real STV elections, parties sometimes hand how "how to
vote" cards. I like Asset because it is not faction-dependent. The
voters simply choose whom they want to represent them and this person
then either represents them or helps choose who will.
There are certain possible vulnerabilities with asset voting that
don't exist with large-scale elections; the problem with asset is its
very strength. You can vote for yourself if you want in Asset, and
then you become an elector who can participate in the direct process
of composing a Council. However, you would then be just one person,
and in situations where security is a problem, and police protection
must be assigned, it's not practical to protect every single isolated
person. But my guess is that these hazards don't exist for you. I
presume that there is a list of party members, and that is as big a
security risk as I could imagine being real.
The other problem is that someone could demand that you vote for
them, and if you don't, and they don't get any votes, they will know
that you didn't. Of course, you could simply lie. "Those idiots, they
can't even count the votes! Or, wait a minute, what did you say your
code was? Damn! I maybe I wrote it wrong. Ah well, next time."
(Because there can be a huge number of "candidates" in Asset, there
would be no candidates' names on the ballot. And that is the norm in
small organizations, it seems you are trying to use
large-organization process! I.e, printed ballots with the names of
nominated candidates, which then favors them, computer processing of
ballots, etc. Why?)
Usually the negotiating goes on until shortly before the voting, so
I am not sure if the added value would be so big. Normally the party
fractions have this preference ordering set up anyway.
The former system breaks the principle of the secret ballot.
How? Asset uses secret ballot. I mentioned a quite minor problem at
the edge. If someone tried to use major coercion, as soon as they had
two votes, they would be unable to tell who voted for them and who
didn't. (Presumably they voted for themselves, so if they get one
more vote, they can't tell who it was, it is only if they get *no*
other votes that they could tell their victim did not comply. Maybe.
Errors are made in canvassing, happens all the time, I've seen it happen.
There is no way that vote coercion could cause a systemic problem and
affect election outcomes that couldn't happen with any voting system.
People think, sometimes, that the electors could be bribed or
coerced, after all, they are public voters. But this would apply to
any voting that takes place in the open, at meetings. And it is going
to apply most of all to the actual members of your council. Where it
would hurt. Bribe an elector who doesn't sit on the council, you get
anything from 1 to a few votes. Bribe a council member, you get a
whole vote on the council.
That's not the reason not to use Asset. It's an excuse.
----
Election-Methods mailing list - see http://electorama.com/em for list info
