2016 election results using alternative models

Published in

From the Desk of Gregory Varnum

3 min readDec 19, 2016

The results of this year’s election have resulted in yet another national discussion around how we elect our President and Vice President.

Currently, most states use a “winner take all” setup for the Electoral College. There are a few alternative proposals to this. I have chosen to break down the numbers by the two that I believe have the best chance of getting national acceptance (based on recent proposed changes that actually made it to ballots).

I ran some numbers from this election up against these two methods, along with an additional variation on each, and come up with these numbers:

Here is some additional explanation on what is depicted in this graphic.

Pennsylvania Method

The first — which I am referring to as the Pennsylvania Method — was proposed (and ultimately rejected) in Pennsylvania in February 2013. Essentially, it takes all but two electoral votes, and divides them among the candidates based on the state’s popular vote. It then awards two electoral votes to the candidate who received the highest support in the popular vote.

Here is an example using this method in California, which with 55 electoral votes gives us a larger number to show impact with third party candidates better.

Candidate A

49% of popular vote
28 electoral votes

Candidate B

46% of popular vote
24 electoral votes

Candidate C

5% of popular vote
3 electoral votes

Colorado Method

The second — which I am referring to as the Colorado Method — was proposed (and ultimately rejected) in Colorado in 2004. It divides the state’s electoral votes proportionately based on the popular vote.

Here is an example using this method, again in California with 55 electoral votes.

Candidate A

49% of popular vote
27 electoral votes

Candidate B

46% of popular vote
25 electoral votes

Candidate C

5% of popular vote
3 electoral votes

Plurality vs. majority

As the numbers for 2016 show, a strict interpretation of these methods does not provide any candidate with a majority (270) of electoral votes, even though it provides them with a plurality. Meaning than they got the most, but not 50% or more. Currently, the US Constitution requires a majority, so this would prompt a deciding vote by the US House of Representatives, which is not ideal either.

There are some possible ways to resolve this. One is to amend the Constitution (which would likely already take place to implement all of this anyway) to require a plurality and not a majority. Another possible solution is to require a minimum percentage of popular votes before a candidate can be eligible for electoral votes. That would not reduce the possibility of a US House of Representatives intervention as much as switching to plurality, but would greatly reduce the chances over a stricter application of the models.

To help account for this possibility, the above graphic also has columns for a 5% minimum. The 5% in general is arbitrary, but seems logical and reasonable. Overall, this does not change the outcome unless the minimum to win is 270.

In case you are curious, if there is a tie or using a majority setup, if any candidate does not get 270, the decision goes to the US House of Representatives (as I mention above). There, each state delegation gets one vote, and they select from the top 3 candidates.

References and further reading

Photograph attributes

Donald Trump August 19, 2015 3 by 2 by Michael Vadon
Hillary Clinton by Gage Skidmore 2
Gary Johnson campaign portrait by Office of Governor Gary Johnson
Jill Stein by Gage Skidmore
Evan McMullin 2016–10–21 headshot by Remember the dot