[EM] 12/22/02 - Markus Schulze Wrote and Wrote again:

[James Gilmour]

Markus wrote:
> In Australia, the voter has either to cast an X-vote for one and
> only one party or to rank all candidates. I suggest that a voter
> should be able to rank parties and candidates

I would not make any provision for ranking parties.  The voters should rank the
candidates, from "1" to as many or as few as they wish.  The "ranking" of the
parties should be an outcome, not an objective..

> and that the used
> STV method should be able to handle equal rankings.

This is not easy (not possible with paper ballots sorted and counted by hand), but
why stop at equal rankings?  Once you start down this road you may as well allow
the voters to assign the "weights" to their rankings.  B L Meek, who devised the
Meek version of STV-PR, described a procedure to take account of what he called
"intensity of preference".  He did this as a mathematical demonstration that
individual preferences could be fairly aggregated while still taking intensity of
preference into account.  He stated that he did not consider it a practical method
and he did not advocate its adoption.

His paper describing this method was reproduced in Voting Matters, Issue 6, May
1996, pp 1-4.  The paper was first published (in English) in Mathematiques et
Sciences Humaines, vol 13, No 50, 1975, pp23-29.


> By the way: Party boxes are not a requirement for large districts.
> The Cork City Council was elected in a single 21-seat district --
> without party boxes. And in 1925, the Irish Senate was elected
> in a single 19-seat district -- again without party boxes.

All I would add to this is that "party boxes" are completely and utterly
unnecessary, no matter how large the district.  They are a recent perversion and
have no place in STV-PR.

James

----
For more information about this list (subscribe, unsubscribe, FAQ, etc), 
please see http://www.eskimo.com/~robla/em