[EM] electoral college/Serious thoughts

[Curt Siffert]

Under your plurality suggestion, Bush would have won in 2000 even 
though Gore was the Condorcet Winner.  Details here:
http://www.museworld.com/archives/001177.html
but in short, the calculations result with:

Bush: 48.17%
Gore: 48.01%
Nader: 2.77%
Other: 1.05%

of the Electoral Votes, if you allow fractional EVs.  Nader spoils Gore 
in this case, too.  Also note that no one got 50% of the EVs, so no one 
got 270 - it would have gone to the House.

There are a lot of bizarre problems if you don't allow fractional EVs, 
for instance a swing state with an even number of EVs would not have a 
lot of campaigning invested in it, while a swing state with an odd 
number would.

The problem with it going to the house is that gerrymandering means 
that the GOP has the advantage there even if the CW is Democrat.

Curt

On May 1, 2004, at 9:47 PM, Dave Ketchum wrote:

> On Sat, 1 May 2004 19:55:20 -0700 Curt Siffert wrote:
>
>> Dave -
>> I had a little trouble understanding what you wrote in your original 
>> email.  But I don't think you have it wrong... I think.
>> I don't think awarding by congressional district is a good idea, due 
>> to gerrymandering, and due to the scenario you describe (10 A 
>> electors).
>> I do think the best idea is to award EVs on a state level, 
>> proportionally, depending on the popular vote in that state.  
>> Assuming I interpret your view correctly here, I agree with you.  At 
>> least, I agree theoretically.  The problem is that I don't see how to 
>> implement it fairly.
>> I think there is a problem in determining the percentage order of 
>> finishers in a state, though.  To proportionally award EVs, you need 
>> to be able to communicate the winners in a form relative to 100%:
>> 49%  Bush
>> 48%  Kerry
>> 3%   Nader
>> You can do this with plurality, by only counting up first place 
>> votes, but you get right back to the spoiler problems.  Since IRV 
>> focuses on first-place votes, it has the same problem.
>
> I was saying count the votes purely by Plurality rules (as is done 
> now), and then award electors proportionally.  Spoiler problem exists, 
> but IS NOT so serious as in current winner take all.
>
> Looking at the above votes, Bush and Kerry each get about half the 
> state's electors and in big-enough states, such as NY, Nader gets 
> about one.
>
> IRV and Condorcet are oriented to winner-take-all - which is exactly 
> what I wish to escape.
>
>
>> And Condorcet is not reliably communicable in this way.  How would 
>> you approximate 100% worth of consensus, and communicate that Nader 
>> got x% of that consensus, Kerry got x% of that consensus, etc?
>> In this example:
>> A->B: 60-40
>> A->C: 51-49
>> B->C: 51-49
>> A is the CW.  If A is ignored, B beats C.   However, C is a lot 
>> closer to winning than B is, in terms of vote-switching cost.  If C 
>> was given less EVs than B, C would have a very convincing political 
>> argument to make in protest.
>> A few threads ago we talked about what sort of math to use to figure 
>> this out - I'm still going through them to see how they fit a variety 
>> of scenarios, but so far I'm not sure any of those counting ideas 
>> we've come up with so far are robust enough for a presidential 
>> election.
>> So, due to my inability to see a clearly fair way to determine how to 
>> award proportional strength, under any vote-counting system, I am 
>> having trouble advocating proportional EV.
>> (The other problem is we'd have have to remove the thing that says if 
>> anyone gets less than 270 EVs, it would go the House - this approach 
>> makes the scenario much more likely.)
>
> Disagreed.  I anticipated Nader getting some electors - but so few 
> there is no chance of his becoming President.  Those who nominate 
> these electors could, as part of the nomination process, specify how 
> they shall act if Nader is not about to win (some states have laws on 
> this topic to consider obeying, or getting rewritten).
>
> ALSO, how is getting the House involved on ties a disaster?
>
>
>> Curt
>
> -- 
>  davek at clarityconnect.com    people.clarityconnect.com/webpages3/davek
>  Dave Ketchum   108 Halstead Ave, Owego, NY  13827-1708   607-687-5026
>            Do to no one what you would not want done to you.
>                  If you want peace, work for justice.
>
> ----
> Election-methods mailing list - see http://electorama.com/em for list 
> info
>