Yanni Cooper [mailto:[EMAIL PROTECTED]] wrote:
> 
> [EMAIL PROTECTED] wrote:
> > 
> > The system of voting used in the United States is known as a
> > plurality vote, meaning that each voter gets one vote, and the
> > candidate with the most votes wins.  
<snip/>
 
> No no, in the usa we have what is called an electoral
> college... our representatives vote for us... 

Hmmm, I wonder what happens when you map the Borda vote (for example)
over an electoral college?

For example, the colleges could vote in a secondary Borda vote with the
order specified by the populace:

Here's a contrived example with nine voters (abcdefghi), and three
candidates (ABC) - and nearly all of the voters are polarized either
A>C>B or B>C>A. Plurality suggests it's a close race between A and B,
third-party votes for C deciding the call (e.g. a hypothetical Gore vs.
Bush vs. McCain runoff). Borda suggests a compromise on C to minimize
unhappiness with the final choice.

a: Plurality: A   Borda: A-3, B-1, C-2
b: Plurality: C   Borda: A-1, B-2, C-3
c: Plurality: B   Borda: A-1, B-3, C-2
d: Plurality: A   Borda: A-3, B-1, C-2
e: Plurality: B   Borda: A-1, B-3, C-2
f: Plurality: B   Borda: A-1, B-3, C-2
g: Plurality: A   Borda: A-3, B-1, C-2
h: Plurality: B   Borda: A-1, B-3, C-2
i: Plurality: C   Borda: A-2, B-1, C-3

The popular votes come out:

Plurality:  A: 3,  *B: 4,   C: 2
Borda:      A: 16,  B: 18, *C: 20

So plurality winner: B, Borda winner: C.

But breaking into three colleges, 1: abc, 2: def, 3: ghi:

1: Plurality: (3-way tie)
   Borda:     A: 5,  B: 6, *C: 7, so votes: A-1, B-2, C-3
2: Plurality: B
   Borda:     A: 5, *B: 7,  C: 6, so votes: A-1, B-3, C-2
3: Plurality: (3-way tie)
   Borda:     A: 6,  B: 5, *C: 7, so votes: A-2, B-1, C-3

(Assume the ties cancel or some such.)

Plurality:  A: 0, *B: 1,  C: 0
Borda:      A: 5,  B: 6,  C: 7

So plurality winner: B, Borda winner: C

However, I'm suspicious of the mapping step - how the individual college
tallies are mapped into the ordering they use for the overall vote. I
suspect that this skews the results since the subtle weighting
differences expressed in the Borda tallies become much more extreme
differences.

Note, though, that this is analogous with the current electoral college
system with plurality voting, where the tallies for a college result in
a filtering to a single choice. This is the point of the college system
(and actually allows more individual say in the election - your one vote
has a higher chance of tipping the balance than in a straight popular
election). However, there may be subtle differences that are less
aparent and so such a system should be analyzed.

It would be interesting to see the Borda results for these scenarios -
numbers are pulled out of thin air - in both electoral college and
non-college examples:

        43%: Gore  > Nader > Bush
        45%: Bush  > Gore  > Nader
      12%: Nader > Gore  > Bush
           
        48%: Clinton > Bush > Perot
      42%: Bush > Clinton > Perot
      10%: Perot > Bush > Clinton

Just looking at those suggests that (if those imaginary numbers were
correct) implementing a Borda system would let Gore win over Bush -
which is what the vote-trading movements are attempting to accomplish -
but still keep Perot out of the mainstream, as one would expect.
However, this system does nothing on its own to enable a candidate like
McCain who (for whatever reason) had intermediate appeal for most
voters. The party system would eliminate such a candidate when they made
their top pick, but the voting system would allow the potential
candidate the real hope of seizing this popularity and running outside
their home party. How the parties would adapt to this would be very
interesting.

It may be that the current system is actually better off in the long
run. I agree with JDG that the current two party system very
successfully keeps the US in a slow-motion dance around some ideal (but
ever-changing, and locally defined) center. A good 200+ years of history
and understanding of that system ingrained into the people with their
fingers on the ballots is not something to be trifled with. In
simulations a system might look great, but if it confounds voters who
weren't raised on it the consequences might be pretty bad.

Joshua
(Who'll just be glad when the damned thing is all over and the
inevitable 4 years of complaining about the winner can get started.)

Reply via email to