[EM] Any published scholarship discussing or describing the Center Squeeze effect?

[Kristofer Munsterhjelm]

On 08.02.2023 17:02, robert bristow-johnson wrote:
> 
> 
> Subject header says it all.  Do any of you know of or have a
> reference to any decent scholarship discussing the Center Squeeze
> effect that Hare RCV has?
> 
> There is a really stupid page at the Center for Election Science
> describing what they call "Center Squeeze" of FPTP.  But there is
> nothing in it about **center** squeeze.  It's about the problem of vote
> splitting with two clones that happen to be in the center.  There is
> nothing about it which demonstrates any bias against candidates in the
> *center* of the spectrum.  It's just about clones, wherever they lie on
> the spectrum.
> 
> I have always understood (and even put it in my paper) that Center
> Squeeze was is a problem of Hare RCV (IRV) and the source of the problem
> is that Hare RCV is opaque to second-choice votes until such votes are
> promoted to effectively first-choice and actually counted.  The centrist
> candidate is at a disadvantage because the centrist candidate can expect
> to receive a lot more second-choice votes from the Left and Right than
> those two candidates on the Left and Right can expect to receive from
> voters in the opposite wing.
> 
> Anyway, I am looking for a good scholarly reference, if one exists.

The Electowiki page on center squeeze gives three, here: 
https://electowiki.org/wiki/Center_squeeze#References

I couldn't find much otherwise. It seems that the term "center squeeze" 
is not used much in the academic literature. The closest I could get by 
using Google Scholar is these:

ZHANG, Jiahua. The Pathologies of Voting Schemes. 2020: 
http://homepage.divms.uiowa.edu/~jones/voting/4980.2020/projects/zhang.pdf
	This is a term project detailing different types of voting method 
failures, including center squeeze. References RangeVoting and Jameson 
Quinn's VSE stats.

CADDICK, Zachary A. Learning, Choice Consistency, and Individual 
Differences in How People Think Elections Should be Decided. 2023. PhD 
Thesis. University of Pittsburgh. 
http://d-scholarship.pitt.edu/44049/1/2022_12_19%20ZAC%20Dissertation%20Manuscript.pdf
	The dissertation is about who people consider to be the rightful winner 
in hypothetical elections. (Caddick finds that most people point to the 
Plurality winner - perhaps out of familiarity - although they also think 
IRV and cardinal voting are more fair than Plurality.) Page 37 mentions 
center squeeze as a weakness ("quirk") of Plurality and IRV, and page 39 
shows the presentation of center squeeze given to the participants of 
the study.

There's also James Green-Armytage's paper on nomination incentive, where 
simulation results show that IRV produces an incentive to exit: 
candidates can help an aligned faction by leaving. But similar to what 
you say, that's more about cloning in general than it is about center 
squeeze in particular.

I'm not sure where the exact term "center squeeze" comes from, but I 
would guess Jan Kok (who was the first to use it on EM, in 2006) was 
inspired by 1D Yee diagrams (like this one: 
https://commons.wikimedia.org/wiki/File:Voteline_center_squeeze.png?useskin=vector 
); and that the term itself was picked up by Warren Smith, and from 
there by William Poundstone, who wrote "Gaming the Vote" in 2013.

-km