[EM] Re: "Approval Later-no-Harm",

Chris Benham chrisbenham at bigpond.com
Tue May 10 10:00:55 PDT 2005

I  had suggested this as a maybe-interesting criterion which applies to 
hybrid rankings/approval methods:

"Disapproval Later-no-Harm:  Ranking a disapproved candidate must never 
harm an approved candidate."

You responded (Sun.May8):

>Unfortunately, I don't think that can be guaranteed without violating 
>the Condorcet Criterion (as Ted points out below). Any time you rank 
>another candidate, you take a chance of making that candidate the CW and 
>taking the win away from one of your higher-ranked candidates -- no 
>matter how far down you rank the new candidate. Anyone who insists on 
>LNH must give up on CC, I think (Please correct me if I am wrong about 
You are not wrong regarding the usual "standard" version of  the CC, 
But Later-no-Harm  *is* compatible with  "Condorcet(Gross)", which says 
that a candidate
that is supported by more than half the voters in pairwise comparisons 
with all the other
candidates must win.
If  none of the voters truncate (or equal rank), then the 
Condorcet(Gross) winner will be the
same as the Condorcet(Net) winner; and if the method meets LNHarm then 
the voters will
have no incentive to truncate!
Kevin Venzke was making this point a while ago in his advocacy of  
MinMax(Pairwise Opposition),
aka Woodall's  "Minimum Doubly-Augmented Gross Score", and  of  
"Condorcet (doubly-augmented gross)
top tier" (CDTT) methods.
The CDTT methods can fail LNHarm when there are more than three 
candidates in a majority-strength
top cycle, but unlike MMPO they meet  (mutual)Majority and (votes-only) 
Minimal Defense.


Probably my favourite plain rankings method  (certainly one of my top 3) 

I am of the view that it is possible and desirable to have the situation 
where the big majority of  voters
are innocent of  strategy and/or not interested in strategy, and  I 
strongly believe that the "official assumption"
should be that there is no strategy problem and no problem in 
determining the will of the voters from their votes.
So  I'm  completely against relying on any random process like random 
ballot (except as a last resort when
candidates are tied according to all the reasonable deterministic 
processes), or on explicit strategy devices
(like ATLO).   If   the voting method  has  a big role for   Random  
Ballot,   the  implied  official  assumption
is  "No voting method  is immune to strategy, and we know that you 
voters can't be trusted!".

CDTT,IRV is  the only CDTT method without a clone problem that meets the 
"Weak Burial Resistance"
"If  x is the CW (and wins), and on more than 1/3 of the ballots ranked  
above y and z; and afterwards on
some of the ballots that rank y above x and x not below z, z's ranking 
relative to x is raised while
keeping y ranked above them both, then if there is a new winner it 
cannot be y."


That criterion is incompatible with Mono-raise, but I consider the 
exchange a huge bargain.
(Especially in public debate versus FPP and plain IRV)  I  don't see how 
we  can wear this  James G-A
Burying example:
46: A>B
44: B>C  (sincere is B>A)
05: C>A
05: C>B
A is the sincere CW.
First-preference supporters of  a sincere CW with more than 40% of the 
FPs  simply shouldn't have to
bother with "defensive strategy"!

Chris  Benham

More information about the Election-Methods mailing list