[EM] In defense of the Electoral College (was Re: Making a Bad Thing Worse)

[Kristofer Munsterhjelm]

Raph Frank wrote:
> On Thu, Nov 6, 2008 at 2:58 PM, Steve Eppley <SEppley at alumni.caltech.edu> wrote:

>> For recounting in close states to affect the outcome, the leader's share of
>> the EC (prior to recounts) would need to be very very close to half of the
>> EC.
> 
> If a State has 10 seats, then it would be 0.2% per seat.
> 
> However, I would agree, in most cases, there wouldn't be an issue, as
> it would require 2 things to happen at once.  First, there would need
> to be an extremely close national election and also an extremely close
> State vote.

Perhaps the chances of this could be estimated by looking at how past 
presidential elections have gone. The national data is easily 
accessible, but I don't know about the state data; I know Wikipedia has 
diagrams of which way the various states voted in each presidential 
election, but to my knowledge, that doesn't show the margins that you'd 
need to see which states were close.