For all the talk about how Trump and other Republicans are kneecapping themselves with non-white voters, especially with Hispanics, Democrats are apparently actually surprisingly NOT dependent on minority votes to win the presidency.
For the scenarios in this post, I am using the 538 demographic calculator and leaving everything at the 2012 baseline, except for changing minority turnout as I describe.
First, what if literally ZERO Hispanics or Asians voted? I know that nothing like that is going to happen, but the point is to see how dependent Democrats are on votes from Hispanics and Asians in order to win. So with Hispanic and Asian turnout both dropping all the way to ZERO,
here's what you get:
Clinton 283 EV; 48.1% PV
Trump 255 EV; 50.2% PVNV, CO, NM, and FL flip. Trump wins the popular vote 50.2%-48.1%, but loses the electoral college 255-283. Even while winning the popular vote by 2 points, Trump is still a ways away from winning the electoral college. The next closest states are Ohio (50.1%-48.3% Dem) and Pennsylvania (50.1%-48.6% Dem).
What's particularly interesting about this is that Clinton wins the electoral college comfortably, even though Trump wins the popular vote by more than 2%. Clinton could lose one other state like Ohio, Pennsylvania, or Virginia, and still win.
So next, if we keep Hispanic and Asian turnout at ZERO, how far does black turnout have to drop in order for Trump to win?
Black turnout has to drop to 58%, at which point VA and PA flip to Trump:
Trump 288 EV; 51.0% PV
Clinton 250 EV; 47.3% PVPA flips before OH because it has more Hispanics and Blacks than Ohio. Democrats are more reliant on minority voters to win PA than OH.
So the Democrats' electoral college advantage and lack of dependence on Hispanic and Asian voters is such that Trump could win the popular vote by almost 4%, with ZERO Hispanic or Asian voters turning up to vote, and with black turnout dropping 8 points to 58%, and Trump would only BARELY just win.
What this shows pretty clearly is that the Democratic electoral college majority does not depend much at all on Hispanic and Asian voters. Support from Hispanic and Asian voters is just gravy.
But what about African Americans? How much more reliant are Democrats on African Americans than upon Hispanics and Asians to win the electoral college majority? In this scenario, I keep Hispanic and Asian turnout at the same levels as in 2012, and instead see how far Black turnout could hypothetically drop, with Hillary still winning. Again, I know that this won't happen. The point is to see how reliant Democrats are on black voters for an electoral college majority.
With this scenario, black turnout drops all the way from 66% to 30%:Clinton 272 EV; 48.4% PV
Trump 266 EV; 49.8% PV
FL, VA, and OH flip. Trump wins the popular vote 49.8%-48.4%, but loses the electoral college 266-272. He loses PA 49.3%-49.2%, so if black turnout dropped to 29% instead, Trump would narrowly win with PA as the tipping point state.
So the conclusion from this is that Democrats are more reliant on the Black vote than the Asian vote and the Hispanic vote to get an electoral college majority. But still, in a scenario with incredibly low black turnout, in which Trump wins the popular vote by 1.4%, he still loses the electoral college. Overall, Democrats are surprisingly NOT dependent on minority votes to win. When you drastically cut minority turnout, Trump could win the popular vote by a substantial margin but still lose the electoral college.