[EM] Fwd: Duncan Proposal Draft

[Kristofer Munsterhjelm]

On 10/13/23 20:25, Toby Pereira wrote:
> I also like proportional representation, but there are many different 
> elections for many different things, and there will always be a need for 
> single-winner methods. Because of that I'm not sure it's necessary to 
> make the same point in every discussion about single-winner methods, 
> especially specific discussions about solving specific problems (e.g. 
> burial) rather than more general discussions.

I imagine you could have cycles in PR too. Perhaps something like having 
a party list case of a Condorcet paradox, with the number of seats not 
being a multiple of the number of parties, e.g.

31: A1>...>An>B1>...>Bn>C1>...>Cn
30: B1>...>Bn>C1>...>Cn>A1>...>An
29: C1>...>Cn>A1>...>An>B1>...>Bn

with some prime number of seats to be filled.

That nobody has found an obvious extension of the concept of a Condorcet 
winner to a multiwinner context doesn't need to imply that PR in itself 
will make the problem disappear.

A better candidate for dissolving electoral problems entirely would be 
asset -- or sortition.

By the way, although I've got Richard plonked,

>> Time to move on from the ancient Greek conception of democracy, as
>> to elect a tyrant, unconditionally

seems wrong. According to Wikipedia, the Greeks (the Athenians at least) 
filled most of their offices by lot. Elections were limited to financial 
officials and generals.

-km