On 8.7.2012, at 12.54, Michael Ossipoff wrote: > On Sat, Jul 7, 2012 at 4:29 PM, Juho Laatu <[email protected]> wrote: > On 7.7.2012, at 21.04, Michael Ossipoff wrote: > >>> But your concern probably is that a party could deviously ask a candidate >>> that they like, and who is, for all intents and purposes, a party candidate >>> of theirs, to run as an independent, with no official party designation, >>> and no mention of a party connection, by hir or the party. >> >> Yes, that's my concern. Except that I expect most party P voters to know >> very well that this candidate that pretends to be independent actually is >> set by party P. Most voters of this candidate would thus be supporters of >> party P. (And those voters would vote for party P in the national vote.) >> >> I answered that concern. It's a concern that could be raised in regard to >> any topping-up ("additional-member") system. And it's a concern that is >> easily answered, as I answered it. >> >> Every party P voter who nationally votes for the party independent doesn't >> vote for P. If non-P voters vote for the party independent, it's because >> s/he has appeal apart from P-ness.S/he deserves those votes therefore. So >> what's the problem? > > In the strategic scenario the idea was not that the voters vould "nationally > vote for the party independent". Their national vote would be given to party > P. Only the local vote would be for the "party independent" candidate (that > is mentally a party P supporter, although has been listed formally as an > independent candidate). Would't this lead to a working strategy as I > described it? > > [endquote] > > Ok, now I see the problem. But isn't it a problem of all topping-up systems? > How do they avoid it in countries that use topping-up systems?
The most straight forward approach in avoiding this problem is maybe to tie the local and national votes together e.g. by deriving the party from the local vote. In the STV based version one could derive the party from the first candidate of the ranked vote, or maybe fractionally in proportion to how each vote will support different candidates. > How do they avoid it in Germany, for example? I'm not an expert on this. It seems that this type of strategies are not a problem in Germany. Maybe others know better and will comment. http://en.wikipedia.org/wiki/Mixed_member_proportional_representation http://en.wikipedia.org/wiki/Electoral_system_of_Germany > > Yeah, that hadn't occurred to me. It's too bad, because it would be nice to > have the local districts that some people consider important, and also the > better proportionality possible with national PR. Myself, I consider the > policy platforms more important than local districts. > > So, if that problem is really a problem (and maybe it is), then I'd rather > have national PR than district PR, if it isn't possible to fairly have both. I think one can achieve quite good national PR, district PR (number of seats per district must be large enough), geographic proportionality (even representation from all areas) and local prepresentation at the same time. There is however a balance between different requirements. For example with single-seat districts + top-up seats one must give up proportionality withing the districts. Juho > > Mike Ossipoff > > > Juho > > > > > > ---- > Election-Methods mailing list - see http://electorama.com/em for list info > >
---- Election-Methods mailing list - see http://electorama.com/em for list info
