[EM] Does Droop give the same results as Hare?

[James Gilmour]

Donald wrote (in part):
>   Second:  Your example tells us that you are a man of many 
> hats.  Today you are wearing your `Political Party' hat.  
> Most days you wear your `Independent Candidate' hat (to show 
> that you are a champion of the independent candidate).

This analysis might suit Donald's purpose, but it is a gross misrepresentation of anything I have
ever written.  I prefer candidate-based voting systems over party-based voting systems because I do
not want to see the power of the parties increased at the expense of the voters.  I carry no
particular torch for independent candidates.  I regard political parties as essential components of
the democratic system, but I favour voting systems that treat all candidates similarly.
 
>   Third:  Your example tells us that Droop will benefit the 
> political parties by averaging votes to protect party 
> proportionality.  I already knew that and I approve of 
> protecting party proportionality in partisan elections, but 
> only in partisan elections (my new elimination rule should 
> replace the Droop quota, but that's another discussion).
> 
> In any non-partisan STV election the Droop quota should not 
> be used because every candidate is expected to stand as an 
> independent candidate.  So, if we were to make a second 
> example by changing your example to a non-partisan election 
> then the quota to use should be the Hare quota and of course 
> the results will be different than your partisan Droop results.

Your distinction between partisan and non-partisan elections is artificial and misleading in
relation to the analysis of the respective effects of the Hare and Droop quotas in STV-PR.  The real
distinction is between those voting systems that are designed to give only (or primarily) PR of
predefined political parties and those voting systems that can give PR of whatever the voters choose
by their responses to the candidates.

The presence of party labels is convenient for the analysis of proportionality, and in the example
we made the further convenient assumption that party was the only factor in voter motivation.   With
the Droop quota the majority opinion group obtained the majority of seats, but not with the Hare
quota.  It is irrelevant that the candidates had party labels.  What matters is the opinion groups
the voters created by the process of voting.  In non-partisan election these opinion groups can only
rarely be identified by convenient labels like "political party", but they exist nonetheless.  The
voters define these opinion groups through their collective responses to the candidates.  It thus
seems reasonable that if a majority opinion group exists that it should obtain a majority of the
seats, or at least, that the voting system should not create the perverse result of giving the
majority of seats to the minority opinion group.

> By the way, if we were to work your example as Party List the 
> results would be three for Party A and two for Party B (same 
> results as your Droop results).

This is true, using either the d'Hondt formula or the Sante-Lague formula.  But it is irrelevant  -
we were looking at an STV-PR election with transferable preferential votes.
 
> You could brag about Droop by saying that Droop STV will 
> elect the same candidates as Party List (Ha Ha).

I don't brag.  As I've said before, you should keep your cheap jibes to yourself.  They do nothing
to advance your argument.

James