I wonder what it would look like if you took the percentages that voted for each candidate and scaled it to the population of the state, and then used that to total the scaled popular vote?
Of course, there's many problems with that, foremost being that you can't assume that those that didn't vote did so in the same relative percentages of support that those that did vote. For example, I imagine there's a higher percentage of Democrats/Clinton supporters in CA and NY that didn't vote compared to the alternatives, and the opposite is likely true of predominantly red states.
Of course, there's many problems with that, foremost being that you can't assume that those that didn't vote did so in the same relative percentages of support that those that did vote. For example, I imagine there's a higher percentage of Democrats/Clinton supporters in CA and NY that didn't vote compared to the alternatives, and the opposite is likely true of predominantly red states.