[EM] ignoring "strength of opinion"

Abd ul-Rahman Lomax abd at lomaxdesign.com
Fri Dec 2 20:58:34 PST 2005


At 07:26 PM 12/1/2005, rob brown wrote:
>Plurality is bad, but not voting strategically in a plurality system 
>makes it even worse, in my opinion.

I think Mr. Brown did not understand what was written. Jan had 
indicated that he had no preference between the leading candidates. 
If this was true -- or at least substantially true -- then he has two 
rational courses of action: vote for a minor candidate whom he *does* 
prefer, or don't vote at all.




>I recommend Range Voting or Approval Voting for primary elections,
>because of their extreme simplicity/ease of understanding/ease of
>manually tallying the votes if necessary, and their tendency to pick a
>winner who has broad support.  (Contrast that with plurality or
>plurality-with-runoff, which can elect fanatical or special-interest
>candidates who have strong core support, but who are strongly disliked
>by most of the rest of the voters.  See
><http://en.wikipedia.org/wiki/French_presidential_election%2C_2002>http://en.wikipedia.org/wiki/French_presidential_election%2C_2002 
>for
>an example of this.)  Range Voting has an advantage over Approval
>Voting in that there are a couple ways that voters can reduce the
>strength of their votes _if_they_wish_ (it's entirely voluntary).
>
>
>Well I guess I just disagree.  Reducing the strength of your vote, 
>in my opinion, is a downright silly option to allow.

Mr. Kok first said that he preferred Range or Approval. Approval does 
not involve weak votes, except that Approval involves abstention in 
the pairwise election between two candidates both receiving one's 
vote. It is the voter's choice whether or not to do that.

Approval is a Range method, and Range *does* allow the casting of 
weak votes.  Why would it be silly to do that, if, indeed, one's 
preference is weak?

I followed this to the extreme in the last local election. There was 
a tax override on the ballot, and I did not consider myself 
sufficiently informed to vote either for or against it. So I 
abstained. That is the ultimate weak vote: it fully leaves the matter 
to the other voters. In Range, one can cast a vote which is strong, a 
full vote, or one may cast a vote indicating a weak preference, and 
it will be counted that way. One may vote Range as if it were 
Approval, using the same recommended strategy, i.e., vote the maximum 
for the favorite among the front-runners and any candidate one 
prefers to that front-runner, and zero for the rest. Or one may cast 
intermediate, essentially fractional, votes.

>   Yes range is probably better than plurality, but I don't think it 
> solves the problems of plurality nearly to the degree that condorcet does.

Condorcet is a preference-based system. Approval is a 
widest-acceptability system. Condorcet methods can possibly elect a 
candidate who is considered acceptable by only a minority of the 
population, Approval and Range systems would be less likely to do this.

34: A>B>>C
33: B>>A>C
33: C>>A>B

A is clearly the Condorcet winner. But in an Approval election, A 
would get 34 votes, B would get 67 votes, and C would get 33. Which is better?

Well, that might depend on the situation, but *generally*, I would 
suggest, the most democratic government is the government which 
enjoys the *consent* of the largest number of citizens. Approval 
potentially maximizes consent, approving a candidate being tantamount 
to consenting to being governed by that candidate (in governmental elections).

>   I don't like approval for similar reasons that I don't like 
> range....I think both force the voter to think strategically, while 
> condorcet almost completely removes strategy from the equation.

At a cost. A very significant cost.

Approval is one system where there is never any incentive to give 
your favorite less than your maximum vote.

Attempting to protect voters from voting strategically is attempting 
to protect them from thinking about the real-world effects of their votes.

But, in fact, I'd do away with elections, as we generally understand 
them, entirely. Elections are *not* intrinsic to democracy, at least 
not where representation is concerned. The very fact that elections 
have losers is a crucial defect. By "loser" I'm not simply referring 
to the candidate who lost, but to all those who voted for that 
candidate as well. Who represents them?

I've proposed an alternative here many times, and it is my own major 
project. Essentially, it is delegable proxy, which has also been 
called liquid democracy, though I think of it in more general terms 
than the LD advocates generally do. Proxy representation is 
representation by choice, not be election. That is, each individual 
"elects" their own representative. By making such elections 
delegable, or transitive (as in Asset voting, incidentally), it 
becomes possible to create a representative assembly without losers.

While delegable proxy could be implemented as a system with a quota 
to reach to be elected to the assembly, I see no need for that; the 
problem of scale in democracy, the problem of direct democracy 
producing unwieldy assemblies is a result of keeping together full 
participation rights (right to enter motions, argue motions, etc.) 
and voting rights. It is quite possible to separate them, where every 
member (citizen) has full voting rights, but only those reaching a 
quota have full participation rights. If you don't have the quota, 
you can still address the assembly, but only with the assembly's 
permission, or, alternatively, without that permission, through any 
member of the assembly willing to represent your views.

There is actually an implementation plan that begins today and could 
(if the theory is correct, *must*) lead to full practical realization of this.

Proxy democracy is not new at all. All share corporations (as far as 
I am aware) are proxy democracies. Delegable proxy is a new twist, 
though, making it possible for representatives to be chosen on a 
small scale while still creating a relatively small assembly.

And another name which has been given for this -- I think my coinage 
of it was the first -- is fractal democracy, for the delegable proxy 
structure would be a fractal.

http://beyondpolitics.org
http://beyondpolitics.org/wiki




More information about the Election-Methods mailing list