[EM] Posting #2: intro, a plea, LWV, organizing v. IRV, terms & taxonomy

[Joe Weinstein]

Posting #2 - 17 Nov 2000.
(1) Intro. (2) A plea. (3) LWV. (4) Org. vs. IRV. (5)  Terminology and 
Taxonomy.  (6) Glossary

	(1)  Introduction

My first-ever posting,  3 Nov. 2000, gave arguments for high- resolution 
grading.  Thanks - notably to Bart I. and Mike O. - for the responses.  
These contend mainly that hi-res can be simplified to crudest-res, i.e. 
pass-fail (‘approval'), considering strategic voting.  A later posting will 
argue otherwise, and will examine issues of sincerity and strategy.

On Fri. 3 Nov - hours after Posting #1, my wife and I left for a glorious 
getaway week, including for me on 6 Nov. a day of solo dayhiking Grand 
Canyon, S. rim to river and return.

We had already (on Halloween) voted absentee in our Los Angeles County 
version of Election USA 2000, on a slick new Touchscreen system.  Physically 
very nice, but of course just another implementation of the humiliating 
affront to civil rights which is embodied in the usual lone-mark 
ballot-marking requirement.  (What I call ‘lone-mark' is usually called - 
quite undescriptively and misleadingly - ‘plurality'. See Terminology and 
Taxonomy discussion below.)

I return to find a flurry of new postings.  Bart's and Mike's postings re 
LWV and re organizing v. IRV raise real questions: see below.

	(2)  A Plea

Most other recent postings discuss angels-on-pinheads subtleties.  Maybe 
these deserve heed, but I can't readily track acronyms and definitions, and 
it's cumbersome to digress to various (and uncited) web sites.  Please, 
learned authors, let each of your more abstruse postings include a (brief I 
hope) glossary of the used acronyms and trickier definitions (as I have 
tried to do below).   Do this, if only to humor new guys like me, each with 
our respective deficiencies (which in my case include a math Ph.D. and some 
pubs. in logic and in graph theory).

	(3)  LWV

It's hard to imagine that an organization with the cachet of LWV circulates 
position papers only in non-e modes; but, OK, stranger things do happen.  
Bart, please indicate what you would have us Californians do re LWV - 
actions, contacts, etc. - and maybe I can do some of it.

	(4)  Organizing v. IRV

Based on my very recent experience, such organizing should be taken very 
seriously very soon.

I've just joined a Long Beach local civic activist list, to which my first 
posting (2 Nov) concerned the (then-impending) tragedy/travesty of Election 
2000 arising from the lone-mark method forcing people to support just one 
candidate.  The only specific response to my posting was a typical pro-IRV 
tirade, spouting the doubly dubious CVD cult equations:  reform = p.r. = 
IRV.

The cult is spreading because, using its very name and a website, CVD 
conveys the impression that it is an authoritative organization.

I am ready to help create and support a real organization for use of better 
election methods.  To be effective, let's flatter CVD by imitating it and 
then some.  I suggest the following ingredients.

*1.   An authoritative-looking web site. (There is now, for instance, a 
low-key 'Approval' web site, but it doesn't connect to an 
authoritative-looking organization).

*2.   A slick name - objective and authoritative sounding.  Better yet, two 
or even three such names, for the following three different aspects of the 
organization:

	(A)  A small hands-on managing (operating and controlling) organization: 
e.g., ‘Center for...' or ‘Institute for...'.

	(B)  A mass public organization - e.g. ‘Association for ...',  which anyone 
can readily join (e.g. via online checkoff) for no or low cost, just to 
indicate support and receive special info.

	(C)  An information effort whose label implies backup by research, e.g. 
‘Research Program for ...'.  This effort will be operated by (A) and be 
directed at (B) and the broad public.

For instance:
	(A) Center for Better Election Methods
	(B) Association for Better Election Methods
	(C) Research Program for Better Election Methods

For a loftier tone or a broader scope, replace ‘better election methods' by, 
e.g., ‘effective democratic choice', or ‘better election and representation 
methods'.

*3.   An explicitly stated mission to promote superior election methods.  
The mission statement must include attractive and broadly accepted criteria 
for desired methods.  These criteria should include many usual methods, 
while clearly excluding IRV.

	My suggestion: a  qualifying method should be one which ‘faithfully' 
rewards each voter's expression of priorities.  Here, ‘faithfully...' is 
taken to mean or anyhow to imply isotony, alias MONOTONICITY - thereby 
qualifying many non-runoff methods but clearly disqualifying runoff (or 
automated runoff) methods like IRV.

(Face it, true runoff - i.e. potentially multiple ballots for successive 
elimination of options - is inherently non-monotonic.  It seems to me that 
the main purpose of runoff, at least in the 1-winner case, is to guarantee 
that in the last stage the contest can be claimed to be won by a ‘majority' 
- a magical term for some people.)

*4.   Positive promotional efforts for the family of methods which meet our 
criteria - as well as opposition to methods such as IRV which do not.  In 
particular, use of an attractive generic semi-descriptive term or phrase - 
e.g. ‘faithful' - for our favored family of methods.

	(5)  Terminology and Taxonomy

For productive discussions on election methods - let alone successful public 
promotion of superior methods - we would do well to start using for these 
methods a more careful and truly descriptive terminology, and its implied 
taxonomy.

In this spirit, I use the term ‘lone-mark', not ‘plurality', to describe the 
prevalent method.  I also suggest using: ‘pass-fail', not ‘approval'; 
‘tradeoff' or ‘fixed-sum', not ‘cumulative'; and ‘iterated reduction' rather 
than ‘instant runoff' (we can keep the acronym IR here).

Here's more detail.  Consider an electoral contest among various listed (or 
written in) options - candidates or measures.  In the USA, most contests 
call for just one ballot (i.e. voting episode) and just one winner.  For 
this most typical case - and many others - the various usual election 
methods differ primarily in terms of the constraints which each imposes on 
what the VOTER does, i.e. how the voter may MARK the ballot.

Each ballot comprises at least the following three distinct key phases:  
actual voting, i.e. VOTERS MARKING ballots; SCORING, from the marked 
ballots, of each of the options; and DECIDING a winner (or winners) from the 
scores.

	Marking.  Every method has the voter mark a grade - a positive integer or 
zero, directly or coded - for each of one or more of the options.  (An 
unmarked option gets zero by default.)

	Scoring.  At least for the usual case of just one ballot, every usual 
method gets an aggregate score for each option by summing (or if you like, 
averaging) the individual ballot grades for that option.

	Deciding.  Every usual method gets a single winner as the highest-scorer 
(with random means used to decide among tied highest-scorers).

Note that ALL methods usually use ‘PLURALITY' to DECIDE.  It is silly to tag 
a particular method by the noun-adjective ‘plurality': all methods are 
equally ‘plurality'.  Rather, methods differ in quite another way, namely in 
what sets of grades they allow at the MARKING phase.

So, rather than ‘plurality', I describe the prevalent system as  
‘lone-mark': it allows only a single nonzero grade, and indeed just one such 
grade (i.e., 1).

Likewise, ‘pass-fail' better describes the method which allows each option 
to be marked with one of two grades, 1 and 0, i.e. ‘pass' and ‘fail'.  
Again, it's a bit silly to call this method ‘approval': for ALL methods each 
score is intended to register the option's degree of aggregate ‘approval'!

Again, ALL usual methods use cumulation (i.e. summation) to get scores.  
Rather than ‘cumulative' voting, a more descriptive tag would be ‘tradeoff' 
or ‘fixed-sum': summed across the ballot options, the sum of all marked 
grades must be at most a given fixed bound, so that allowed individual 
options' grades are in tradeoff.

‘Instant runoff' too arguably applies to ALL usual methods, both for just 
one ballot and for ‘true' runoff (i.e., potentially requiring more than one 
ballot).  Namely, a one-ballot election may be viewed as a trivial kind of 
runoff, and surely is as ‘instant' as can be.  Moreover, nowadays any true 
runoff method can be automated and also made ‘instant', thanks to computer 
technology.  Namely, at the first ballot, the voters can supply all needed 
contingency information,and then be spared any later actual ballot.  I 
suggest keeping the acronym IR, but taking it to stand for ‘iterated 
reduction', i.e. runoff with one-at-a-time removal of options.

A ‘ranking' method is one where, in the MARKING phase, all nonzero grades 
are required to be distinct and to form a set of consecutive integers, from 
L downward, where L is the number of listed options.  Given that a method IS 
a ranking lone-balloting method, the standard SCORING procedure (by grade 
summation or averaging) amounts to Borda.  So in describing a lone-balloting 
method, the terms ‘Borda' and ‘ranking' should be deemed synonymous.  Only 
for a non-Borda ranking method need the nonstandard scoring procedure be 
specified.

	(6)  Glossary

Abbreviations and acronyms:
	CVD = Center for Voting and Democracy; IRV = ‘instant runoff' voting;  LWV 
= League of Women Voters;  p.r. = proportional representation.

Tricky definitions:
	Monotonicity.  An election method is monotone (more precisely, isotone)  
if, given any election held under the method, the result of changing any one 
marked ballot by raising a winning option's grade (and possibly 
simultaneously lowering the grades of one or more of the other options) must 
again result in victory for that option.

THANKS FOR YOUR HEED.

Best wishes,
Joe Weinstein
Long Beach CA USA


_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.

Share information about yourself, create your own public profile at 
http://profiles.msn.com.