[EM] Re: Stabilizing the electoral college (was Re: electoral college)

[Steve Eppley]

Hi,
Rob B asked:
> Steve Eppley writes:
>> Suppose instead it were winner-takes-all except when 
>> the vote is really close:
>-snip-
>> I've exaggerated because of the limitations of the 
>> text font.  When I say "really close" I'm thinking 
>> about within 1%, or maybe 1/2%.  
>> 
>> This would make recounts less important, since they
>> could only change the outcome by 1 delegate or so, 
>> assuming the initial count isn't attrocious or 
>> fraudulent.
> 
> But recounts could still be important, you've just 
> moved the line....what if it was a difference 
> 0.49999% and the election hung on whether it 
> was possibly really 0.5%?

I'm afraid I don't yet understand Rob's question.  
My proposal doesn't keep the line vertical, it tilts
it so that when the election is close the delegates
aren't awarded winner-takes-all.  For instance, 
if the state's final count is very close to 50-50, 
then each of the two tied candidates would be 
awarded half the state's EC delegates.  

If Rob is saying it's still possible an election
could be so close that a single EC delegate in 
a close state could swing the election, I agree.
But the chance of that would be significantly
reduced, making the system more stable.  For 
instance, in 2000, Florida was close enough 
that Gore and Bush would have split Florida's 
25 delegates, with Bush winning 13 and Gore 12.  
Then Gore would have had a majority in the 
Electoral College. (Neglecting the effect of 
implementing this proposal also in the other 
states that were close in 2000, for which I 
don't recall the details.)  Gore wouldn't have
bothered asking for a recount.  Change the vote 
in Florida enough to swing a few more delegates 
to Bush, and Gore would still have had a majority 
in the EC.  So Bush probably wouldn't have bothered 
asking for a recount either.

The two candidates would still compete mainly in
the close states, since a swing of a small percentage
in a close state would still mean a swing of all 
that state's EC delegates.
* * *
By the way, another idea related to Presidential 
elections, the need for a majority in the EC, and 
the use of a good voting method to minimize spoiling,
is to allow candidates to withdraw from contention
after the (preference order) votes are cast.  For
instance, suppose the Democrats had nominated both
Gore and Bill Bradley in 2000.  Bradley might have
been the winner in New York, since he'd played for
the NY Knicks and had been a Senator from NY, etc.  
But if NY gave its EC delegates to Bradley, this 
could keep Gore from winning a majority in the EC, 
and the election would then be decided, perhaps by 
a partisan vote, in the House of Representatives.  
But Bradley could prevent that if he were permitted 
to withdraw, since then Gore would win NY and 
a majority in the EC.

Of course, the option of "well-timed withdrawal" 
needn't be limited to Presidential elections...

--Steve