Communicative System

[Abd ul-Rahman Lomax]

At 11:04 AM 10/28/2010, Its me Mario wrote:
>I've scanned this document before, given it a bit more attention now 
>but I haven't had the time yet to really analyse it thoroughly.
>
>My own ideas were about a more direct form of democracy, similar to 
>this theory but where the stopping point for votes was a particular 
>outcome rather than an individual who would be entrusted to bring 
>about this outcome. The theory is well written and quite convincing, 
>this form of representative democracy through delegate cascades 
>would certainly be interesting to observe in practice. I think a 
>critical component is the ability for voters/delegates to change 
>their votes at any point in time (this is how the system is kept honest).

Yes.

>I can see this posing a problem for integrating the system with 
>existing forms of representative democracy where individuals are 
>elected for a fixed term.

Fixed term is incompatible with direct democracy. However, it's 
possible to set up a system that minimizes the potential damage.

>  If the most popular candidate within this voting system was to win 
> in a general election they would have a fixed term to operate as 
> they wished, users of the system could change their votes if they 
> didn't like the winner's actions but this winner would still be in 
> office. I think this delegate cascade idea relies so much on the 
> ability to switch votes at will that it would be seriously 
> compromised by any fixed term position.

It is very important to distinguish systems that generate advice from 
systems that control power. "Elections" to office with a fixed term 
express and generate power.

>When I think about integrating a new form of democracy like this 
>with existing forms of representative democracy I think about users 
>agreeing to elect an "administrator" - whose only objective is to 
>faithfully put the plans selected by the voting system into practice.

A common idea but a Bad one, for complex reasons. Decisions should 
not be made by "voting systems," but by deliberative ones. Voting is 
only an aspect of this. I'll describe a possible solution to the 
problem raised.

>  This makes more sense when the outcome of the voting system is a 
> particular course of action rather than an individual who is 
> proposing this particular course of action - but I guess it could 
> be made to work in this way too. The "administrator" could work 
> with whoever the voting system has currently selected until such 
> times as voters change their mind, at which point they would move 
> to facilitating the implementation of policies which the newly 
> preferred candidate stands for.

Voting systems are used to determine results when consensus is 
absent. They involve defined voters. Making decisions based purely on 
popular vote is problematic because most voters will be, in practice, 
inadequately informed. Now, delegable proxy can set up and can be 
used for decision-making, but the devil is in the details, and making 
decisions by voting, even delegated voting, is fraught with hazards. 
Whenever a nexus of power is created, it will attract corruption.

>One other advantage of electing an "administrator" rather than a 
>"candidate" is that it allows voters to turn out to the polls and 
>vote for the system itself, rather than the individual who is at 
>that point in time the system's preferred candidate. To take a 
>simplified example, if 60% of voters/delegates support candidate A 
>and 40% support candidate B at the time of the general election, 
>100% could turn out to vote for an administrator on the 
>understanding that whether candidate A or B actually filled the 
>position would be determined on a week-by-week basis through the LD 
>voting system.

How much power does the administrator have?

>To summarise, its my intuition that one of the strengths of a voting 
>system like this its capacity to separate the making of "policy" 
>decisions from the implementation of these decisions.

That is basically a good idea, in a way. The key is to set up systems 
that advise, that gather and process information in such a way as to 
make the advice maximally trustworthy. The advice flows in two 
directions: it advises decision-makers (people who have been assigned 
authority by whatever process) and it advises the people who choose 
the decision-makers.

>  When people vote in our current forms of representative democracy 
> they make a lot of compromises (we must choose one candidate to 
> represent all of our political views, and this candidate also has 
> to be competent to "put these views into practice").

Yeah, an obvious problem.

Okay, a solution: Asset Voting, used to create a proportional 
representation assembly. The Asset election is held periodically, 
perhaps once a year.

The election for the assembly is single transferable vote, but with a 
tweak and some conditions. Some details I describe may not be essential.

1. The "district" is the entire assembly, and the quota is the number 
of votes divided by the number of seats to be elected. (Hare quota). 
It is possible that some seats will be temporarily vacant, and it is 
also possible, as will be described below, that there may transiently 
be more seats than used for the quota.

2. Any person who receives votes in the Asset election becomes an 
"elector," a public voter.

3. The seats are created when electors put together a quota of votes, 
"spending" them. They may have votes left over, which remain spendable.

4. (Electors may use delegable proxy systems to negotiate vote 
amalgamation, but the actual vote transfers are direct, normally, 
though an elector might possibly name a proxy who can "vote" for 
them, in their absence.)

5. This is optional and voluntary, but when an elector transfers 
votes, they may ordinarily transfer specific votes from specific 
precincts, thus allowing the creation of seats with a regional 
consitituency. Most electors may transfer all their votes to a single 
seat, I expect.

6. The Assembly meets as an normal deliberative body. Only those 
holding seats may enter motions or debate them; however, electors may 
vote. Electors, when they vote, cast a fractional vote according to 
the votes they transferred to seats, and the seat's votes, from them, 
are devalued accordingly. My expectation is that these direct votes 
by electors will normally be unusual and will not normally shift 
outcomes. Electors may vote remotely.

7. Electors may recall a seat, by recalling the votes that they cast 
to create the seat. Such a recall will not necessarily immediately 
cause the seat to leave the assembly, but the voting power of that 
seat might be routinely reduced.

8. In this system, the maximum voting power of a seat is one full 
assembly vote, which represents a quota of votes amalgamated from the 
Asset election.

Most of what I could write about this falls out from the political 
system it would create. Voting in the Asset election doesn't need 
political parties at all. Voters may vote for the person they most 
trust. They may, if they are willing to put in the effort later, vote 
for themselves and participate as electors, i.e., public voters, 
whose votes will be visible.

The system would create a penumbra of electors around the assembly. 
The collection of electors represent the people during the term, but 
they are many. They participate in deliberation only by advising the 
seats. They may vote directly, but normally they will be, on most 
issues, supportive of the seats they elected.

All voters can tell, if the device of assigning voters by precinct is 
followed, who, specifically, represents them in the assembly, and 
they also whom to talk to if they have an issue about the assembly. 
They talk to the elector they voted for. That elector does not know, 
for sure, who voted for him or her, it's secret ballot, but ... the 
electors will normally, I expect, serve small constituencies, and the 
only reason that one might want to talk to an elector other than the 
one voted for would be if one voted for some remote candidate, 
perhaps to get some unusual view represented in the assembly. 
Generally, under a system like this, people will vote for the 
individual they most trust. Themselves, if they must!

Asset Voting was invented by Lewis Carroll (Charles Dodgson) in about 
1884. Brilliant system. He was an early voting systems theorist, and 
he realized a problem with single transferable vote, that most voters 
would only know one candidate they preferred, whereas STV depends on 
voters, ordinarily, voting for a list of candidates in order of 
preference. It is a real problem with actual STV elections that 
ballots become exhausted, that is, that all the votes on a ballot 
have been considered and those candidates eliminated. Further, the 
elimination process is subject to a known hazard, that the best 
candidate by most measures is eliminated because they don't get 
enough first-preference votes, they are a compromise. So STV can 
tend, if used for small numbers of seats in a district, or, worse, if 
used for a single-winner election -- we call that Instant Runoff 
Voting in the U.S. -- lead to the election of the most popular of the 
extremes, opposed by a majority of voters. By allowing candidates to 
transfer votes, no votes are wasted. There is no other system that 
does this, among the well-known proposals.)