[EM] PR/STV Hybrid for multi-winner?

[Adam Tarr]

(If this idea has been suggested before, then I apologize for unintentionally 
copying and throw my support fully behind the original proponent.  If you don't 
care to read my long-winded build-up, just skip to the sixth paragraph, not 
including this one.)

Proportional Representation (PR) has the obvious advantage of matching the 
expressed voter preference as closely as possible in a multi-winner election.  
The problem with PR, in my opinion, is that it limits the voter preference to a 
single party (and a single order within that party) and as such gives too much 
power to the party leadership.

In order for a candidate to advance up the party list, they will be pressured to 
vote in lockstep with the party lines.  Dissenters within a party will have an 
incentive to split off and create an alternative party, which could lead to 
fractionalization and chaotic government, by causing each party to cater to an 
increasingly small constituency (and thereby freezing that constituency out if 
they cannot get elected or join a dominant voting block).  While some PR 
democracies have avoided these problems, others do have these sorts of issues.

STV (in multi-winner elections; I would never advocate its use in single-winner 
elections) solves these problems by allowing voters to single out specific 
candidates that they like within a party.  This allows freer expression of 
preference by the voters, and can motivate a shift in party policies as the 
voters select specific candidates that suit their politics.

While I have never seen a rigorous proof of it, it seems intuitively correct to 
say that as the number of winners in a multi-winner election rises, STV comes 
closer and closer to PR in terms of representing the exact percentage breakdowns 
of the voters.  If a voter's list of candidates never break the quota throughout 
the STV count, then of course his or her vote will not contribute, but the same 
is true for that voter in PR.  But as the number of seats rises, and the length 
of the voter's list rises, the chance that a voters ballot will fully transfer 
rises.  So it would seem that STV, with a large number of winners, offers the 
best of all worlds to the voter: fully proportional representation with the 
ability to select specific candidates.

The problem with such a setup is largely practical: it requires the voter to 
select an extremely long list of potential candidates.  This means the voter 
must either reproduce a long list provided by the parties, or actually go 
through each candidate one at a time and decide the proper order.  Even the 
first option is cumbersome, and is a lot to ask of the voter.  Just ticking off 
all 52 Republican congressional candidates in California, for example, would be 
a huge pain in the butt.

So my idea is this: Implement STV, but allow a _party_ to be entered as a line 
in the ballot.  Entering the party is tantamount to listing the entire slate of 
party candidates, excepting those listed elsewhere on the ballot, in the order 
provided by the party.  In order for a party to be an option for the ballot, it 
must provide a list of candidates equal to the number of seats available.

This seems to solve all the problems.  For example:  

Say I live in Illinois (20 Congressmen) and I want to vote Democratic.  But 
there's one independent candidate I like, and there's a particular Democratic 
candidate, listed tenth on their list, that I like even more.  I also happen to 
hate the third candidate listed with the Democrats.  So I vote my favorite 
Democrat, followed by the independent, followed by the Democratic party, 
followed by the disliked candidate.  This is the same as voting...

1. My favorite Democrat (#10 on the list)
2. The independent
3-4. 1-2 on the Democrat list
5-10. 4-9 on the Democrat list
11-20. 11-20 on the Democrat list
21. 3 on the Democrat list

...in a STV election, but in stead of making 20 or 21 marks I only made four.  
If I am concerned that my vote will not fully transfer (extremely unlikely, but 
possible if the Democrats are extremely unpopular) then I could continue to list 
more candidates or parties, up to the space provided.

If everyone just votes the parties with no changes, it "breaks down" into PR.  
Also note that this works best for larger states; in a one-district state (or 
even a two-district state, probably) it would be far better to use Condorcet or 
Approval.  Fortunately, the states can decide this sort of thing individually, 
so reform movements can advocate STV/PR only in the large states where it will 
work as advertised.

Any thoughts?  Has this sort of method been proposed before?

-Adam