[EM] Re-intro + brief review of some EM pubs

[Joe Weinstein]

Hello, everyone!

This is my first posting in at least seven months, so let me first 
re-introduce myself before briefly reviewing some EM publications that I 
have not yet seen noted on this list.

RE-INTRO.  I am a (Ph.D.) mathematician and now working applied 
statistician.  I started with research in logic and networks, and held 
academic posts, then twenty years ago turned to applied math and stat.  I 
have since worked in energy conservation engineering, then in 
military-related operations research, and for the past ten years in 
environmental conservation and management.

Just over a year ago, our office prepared to relocate, so I found it useful 
to get rid of boxes of accumulated old papers, including math and other 
journals.  On a would-be final read of some of these - notably of an old 
Balinski-Young paper on fair apportionment and of a more recent paper by Don 
Saari on electoral preference paradoxes  - I got hooked on election methods 
and their reform.  My interest has been stoked also by recent and continuing 
involvement in actions for local civic reform and environmental preservation 
in our excessively misgoverned city of Long Beach, CA.

So, about a year ago I joined this EM list, but about seven months ago was 
summarily dropped.   (The EM-list control software claimed that my email 
program was ‘rejecting' too many EM-posts, but I don't know what was 
actually happening.  In truth, I would still prefer to be able to post out 
to the list with my inbox being protected from individual EM-list posts - 
preferably by the EM-list control software rather than my own imperfect 
inbound mail filter.  If anyone knows what method, if any, actually works 
for arranging this, please let me know.)

Anyhow, during the last seven months I've had little time to contribute to 
the list, and have been content to read Yahoo-archived posts. Thanks are due 
to all of you (especially indefatigable Forest and ever-vigilant Bart) for 
the ideas which have made this reading worthwhile.  Just lately I've 
rejoined the list (for which I thank the Moderator), in hopes of being able 
to contribute occasionally.  In particular, I hope soon to post summary 
conclusions which I have gleaned from list discussions.  Some of these 
conclusions confirm my prior beliefs, but others do not.

BRIEF REVIEW OF SOME EM-PUBLICATIONS.  During the year 2000, various pubs 
written by (and often for) applied mathematicians explored the topic of 
election methods.  Perhaps the best, albeit imperfect, was noted prominently 
on this list:  Dana Mackenzie's Discovery Mag article.  In addition, 
however, SIAM (Society for Industrial and Applied Mathematics) presented 
other pubs which I haven't seen noted here.  These were, at first blush, 
quite apropos, as Saari's key article cited above had also a few years 
earlier been published by SIAM, namely, in the mid 1990s in SIAM Review.

As its first title in a projected new series in ‘discrete math', during year 
2000 SIAM put out a monograph, by five Roman academics, under the title 
‘Evaluation and Optimization of Electoral Systems'.  For brevity, I'll cite 
this as IJ -  ‘the Italian Job'.  IJ usefully described various of the 
European electoral systems, and it discussed a broad range of various 
interesting and relevant EM topics, e.g. EM axioms, fair representation 
methods, and criteria for drawing district lines.  However, each of the 
topics was unimpressively treated:  some major approaches were scarcely 
noted, important defects of some noted approaches were not discussed, and 
insightful critiques and suggestions were few.  Despite a plethora of 
formalism, IJ used many imprecise or misleading terms which, as I and others 
have noted here, needlessly plague typical EM terminology and reinforce 
rather than correct for deficient taxonomy.

During fall 2000 (I believe in Oct. or Nov. or both), SIAM News also ran 
several EM-focused articles.  These articles were poorly edited:  in 
particular, they lacked basic cross-referrals to relevant pubs - even to 
SIAM's, including SIAM News itself!  For instance, one article,  about 
Italian electoral politics, never mentioned IJ - despite the very same issue 
of SIAM News running a separate ad for IJ!  Another of the articles, by 
Mackenzie, featured Saari's work - whose main findings had first been 
presented in the above-cited SIAM Review paper - but Mackenzie did not cite 
this paper!  Moreover, despite being targeted at what might be assumed a 
technically more advanced readership, this article of Mackenzie's was 
actually not as rationally organized (nor as generally relevant) as his 
Discovery Mag article.

As a mathematician I was quite disappointed by all these publications.  
Their timing was opportunistic (for USA elections) but they lacked basic and 
readily expressed mathematical insights into our problems of finding and 
confirming good, improved and optimal election methods.

A mathematician who is seriously attacking a given problem will set great 
store on defining (with rationale) just what are the problem's key variables 
(or dimensions), and using these to describe clearly what a solution 
requires.  [This requirement will often take one of three forms: solve for a 
given variable, find a case where the variable is constrained within given 
limits, or find a case which optimizes (maximizes) the variable.]  Key 
variables are also needed in order to define and work with a logical 
taxonomy of problem cases.

Unfortunately, none of these various pubs thus addressed our EM problems, or 
even correctly described them.

Joe Weinstein
Bixby Knolls, Long Beach CA USA


_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp