[EM] (MA-2) Societal institution of communicative action

[Michael Allan]

Raph Frank wrote:
> It might also be worth discussing possible renormalisation of votes
> here to cancel out participation bias.  Also, it might be worth adding
> in actual participation bias in the final election.

OK.
 
> This would mean that if a segment of the population tends to vote,
> then this could be reflected in one of the results.

Yes.  It'll be interesting to compare those results side by side.

> >  The second requirement is that the voters must faithfully carry their
> >  votes over to the principal polls on election day (step 2)...
> 
> Also, this protects against corruption of the system.  If voters vote
> due to intimidation or vote buying, then they can still vote the other
> direction in the final election.

Like a safety valve.  Under pressure it doesn't break.  It fails
gracefully.
 
> OTOH, there is risk of people defecting.  One of the first things that
> the person elected in such a way might do is change the basic election
> system so that people are more likely to vote in the same direction as
> the original communication system.

You mean like gerrymandering?  Exploiting the voting patterns in the
open system?  I guess you're right, that's a vulnerablity.  It might
work in either direction too, for or against the open voters (equally
bad in either case).

> A good system for the final vote might be plurality, but requiring
> that the winner have 50% or more.  If the plurality winner doesn't get
> 50%+, then a new election must be held.
> 
> This encourages pre-election pacts.

I do not consider changes to state electoral systems (where the final
votes occur).  I leave that to you experts...

The medium of assent can never have a "final vote", of course.  People
are always free to change their minds.  So there has to be a kind of
slip-clutch between the lifeworld and the state system (if only for
that reason).

> Have you considered the possibility of using encryption to allow
> anonymous participation.  The Chaum digital cash scheme could be used.
>  Each voter could register for 1 digital 'coin'.  This coin then
> allows them to register a user ID.  (In practice, Chaum's system may
> not be required as there would be no offline trading of the coins and
> also coins would only be handled once, i.e. you withdraw the coin
> using your public name and then deposit it against your anonymous ID).
> 
> The coin might give some basic info about the person -
> age/gender/location.  This would allow renormalisation of the polls
> which keeping identities secret.
> 
> The advantage is thet some people might not want to admit their
> political views publically.  They would then vote differently in the
> secret ballot than in the public communication system, which weakens
> the system.

Or many of them will not bother to participate at all.  That's another
advantage to a secret ballot: higher turnout.
 
> A potential problem would be that it allows direct vote selling as
> votes would be completely transferrable, you could just send your
> 'coin' to another person.  A protection against this would be to
> reissue the coins every so often (say every 3 months).  A sold coin
> would only last that long.

My design strategy (long term) is to allow all methods of
authentication.  Currently I have only the full disclosure (open cast)
method of authentication.  But if anyone wanted to design and code a
PK-encryption method, I think the architecture will be open to it.

My plan is to have a separate, quarantined count for every different
kind of vote (open cast, encrypted coin, etc.).  People can then
choose which count (or combination of counts) they wish to trust.

So we can experiment with different methods of authentication and not
risk the overall credibility of the system.  Any particular method
that proved untrustworthy would simply be devalued by the public -
much like a currency floating on international exchange markets.

-- 
Michael Allan

Toronto, 647-436-4521
http://zelea.com/