Also, am I misunderstanding how you calculated? Why didn’t McMullin get any of Utah’s electoral votes? He got 20% of the vote there.
Yep.
To simplify the calculations, the extract I ran was based on the four major national candidates.
Which meant I missed that one state had a significant fifth candidate.
State: Utah
Votes Cast: 843,353
EC Reps: 6
EC Quota 120,480
Primary Quotas
Clinton 1.97
Johnson 0.23
McMullin 1.46
Stein 0.05
Trump 3.30
After allocation of surpluses
Clinton 2
Johnson 0
McMullin 1
Stein 0
Trump 3
Assuming the OP’s exact calculations are correct, Johnson could elect Clinton by telling his electors to vote for her, or Johnson and Stein together could throw the election to Trump. Neither of those things are happening. But maybe the DNC could dump Clinton and substitute some pro-gun but otherwise socially liberal, fiscally conservative Democrat. If Johnson went for that, at least most voters would have voted for someone who got to exercise some influence over the selection, and I bet more voters would be happy with the result than would have been happy to see either Clinton or Trump elected. If no agreement could be reached, it’s off to the House, just as is done now.
I quite like this system, but I don’t think you can amend the Constitution to get rid of the EC, because it favors small states, and most states are small. And no individual State is likely to adopt this plan, because in most cases it would mean that their votes are split almost down the middle, thus greatly reducing that State’s influence on the outcome.
Actually, under this system you wouldn’t even really need primaries. All of a party’s candidates could run in the general election, with the agreement that they will all assign their electors to the party member winning the most votes. In this case, Sanders supporters and moderate Republicans wouldn’t have had to choose between voting for a candidate they disliked or voting third party. So I’m not sure that this system would actually lead to more third party voting, although I don’t think it would be a particularly bad thing if it did.
No, they still would have to weigh those concerns. I would have voted for Sanders in the general election (and did vote for him in the primary), but I may not have done so under this proposed system as I would not want to even have the possibility of my vote being cast for Clinton. It’s not hard to imagine a Republican thinking the same of his preferred candidate and Trump.
This scenario can be accommodated.
Assume your presidential ballot looks like below:
Clinton
Johnson
McMullin
Rubio
Sanders
Stein
Trump
What you seek could be achieved two ways.
- If preference allocation was mandatory (i.e. you must number every candidate)
In this case …
Clinton 6
Johnson 4
McMullin 2
Rubio 5
Sanders 1
Stein 3
Trump 4
… would ensure that Clinton **could not **get your vote. It would mean however that your vote would probably have ended up in Trump’s numbers.
- If preference allocation is not mandatory (i.e. you need not number every candidate)
In this case …
Clinton -
Johnson 4
McMullin 2
Rubio -
Sanders 1
Stein 3
Trump -
… would ensure that neither Clinton or Trump would get your vote, but if all of your preferenced candidates were eliminated then your vote has been exhausted and isn’t counted in any subsequent round.
Wait, now you’re introducing ranked-order voting to make things even more complicated?
More simply, akennet could have voted Green, so that his/her electors wouldn’t have been pledged to support the Democratic candidate. Of course, it’s possible he or she could still have ended up contributing to the election of some moderate Dem/Green compromise candidate, but it’s highly unlikely that candidate would have been Clinton herself. If even that was unthinkable, he/she could have simply not voted, leaving him/her no worse off than the current system.
Using D’Hondt and a revised vote count (Trump 2,089,062, Clinton 1.877,911, and Johnson 125,293), I also get 8-8. Note that 18 votes would be split 10-8 for Trump, and Johnson would get his first at number 31.
As stated, it all depends on the method used to apportion the surplus votes and we could post duelling polls ad infinitum for no particular advantage. The result is tight,the precise method is critical.
FWLIW the modified D’Hondt simulator here would also give the result as 8:8 [/shrug]
I presume from that statement you have more granular data than the state totals I’m working on, but based on the revised count provided the result is clearer for Clinton to get the last spot.
Clinton 7.80 quotas
Trump 8.68 quotas
Johnson 0.52 quotas
If votes exhaust Clinton gets the last sport by virtue of holding largest proportion
If Johnson votes are reallocated they now need to split 62:38 for Trump to win.
Neither of your scenarios are the same as the one presented by Thing Fish, which is what I was responding to.
Which means the system doesn’t get us away from having to vote for the candidate that we don’t like or voting for third party. What I would have to do is exactly the kind of thing that you said your system would get us out of doing.
Well, if you don’t like the idea of voting for either a major party or a third party candidate, I’m not sure I can help you!
Don’t worry, there’s likely no help for me at all. 
But that was kind of my point. You said that your proposal would mean that a Sanders supporter or moderate Republican wouldn’t have to have to vote for a candidate that they didn’t like or for a third-party candidate, but that’s exactly how it turns out. If my Sanders vote automatically turns into a Clinton vote, then I have voted for a candidate that I don’t like.
No, it’s tantamount to using the popular vote with small state bias.
+1
What you could do, I suspect, is amend the Constitution to allot each state its electoral votes in proportion to the percentage of voters who bother to participate in the presidential election.
To wit: say that the state of Statesota has 28 representatives in the HoR, and two senators, entitling the state to 30 EVs. But only 30% of the eligible voters can be bothered to vote in the presidential election. So Statesota gets 9 EVs in this cycle. Meanwhile, relatively smaller Minnechigan has 18 members in the House, and is entitled to a maximum of 20 EVs. Fully 75% of eligible voters participated, and so Minnechigan contributes 15 electoral votes for the cycle.
A scheme such as this is arguably more equitable, and has the added virtue of incentivizing participation in the electoral process.
I am surprised none of the major news outlets has done the analysis of the 2016 by district to determine the outcome had each state used the same process to award electors that ME and NE used.
EC is never going away. Too many small states would block an amendment. Best option is to get each state to move to a proportional awarding of electors.