Trump v. Clinton is often framed in terms of the lesser-evil principle. I think that’s a valid principle.
However, I think the principle becomes worthless if it never bottoms out. If there’s no threshold below which a candidate, however atrocious, can ever be out-of-bounds, then the comparison is morally compromised beyond recognition.
For instance, some people say not voting for Trump is a vote for Clinton. In a sense.
But suppose you had a choice between Stalin and Mao. Suppose you concluded that Mao was marginally better than Stalin (or vice versa). But when two candidates are as bad as Stalin and Mao (in my hypothetical), is the lesser-evil principle even germane anymore?
What if you said, not voting for Mao is a vote for Stalin. Even if there’s a sense in which that’s the case, so what?