[EM] Papers at the 80th MPSA conference

[Kristofer Munsterhjelm]

On 2023-09-05 18:35, stephane.rouillon wrote:
> 
> 
> Hello,
> 
> It has been a while but I still work on electoral systems.
> Here are the links toward two papers I presented:
> « How to reconcile stability and representation: a weighted crutch option. »
> 80th Annual MPSA Conference 2023 <https://tinyurl.com/2p5yo7pf>
> tinyurl.com <https://tinyurl.com/2p5yo7pf>
> 	
> 
> <https://tinyurl.com/2p5yo7pf>
> 
> 
> <https://tinyurl.com/2p5yo7pf>
> 
> and « Why break Condorcet cycles when we can make them disappear? »
> 
> *https://tinyurl.com/2zuuo2r3* <https://tinyurl.com/2zuuo2r3>

I couldn't find any links to the papers themselves, so I did a little 
bit of digging. In case it may be of interest to others, here they are:

"How to reconcile stability and representation": 
https://www.researchgate.net/publication/370161535_HOW_TO_RECONCILE_STABILITY_AND_REPRESENTATION_A_WEIGHTED_CRUTCH_OPTION

"Why break Condorcet cycles when we can make them disappear?": 
https://www.researchgate.net/profile/Stephane-Rouillon/publication/370167659_WHY_BREAK_CONDORCET_CYCLES_WHEN_WE_CAN_MAKE_THEM_DISAPPEAR/links/6442fff5d749e4340e2b059b/WHY-BREAK-CONDORCET-CYCLES-WHEN-WE-CAN-MAKE-THEM-DISAPPEAR.pdf


On a related note, usually there seems to be little reason to do ranked 
voting with party list PR, because there are so many seats that the 
first preference proportions are accurately represented. But it *could* 
be used to give centrists (additional) kingmaker power, e.g.:

34: A>C>B>D
31: B>C>D>A
30: D>C>B>A
  5: C>A>B>D

If we want to give one of these parties more seats than their exact 
share, it seems more reasonable for that party to be C, not the largest 
first preference vote-getter A. But not here:

34: A>B>D>C
31: B>D>A>C
30: D>B>A>C
  5: C>A>B>D

even though the first preference distribution is identical.

-km