I am hardly a political expert, but I had an idea that concerned voter turnout: what if those eligible voters who didn't vote are counted as being votes for none of the candidates, a sort of "vote of no confidence". This segment of people is usually left out of statistics as an unknown: who knows which candidate they would have voted for? Well, the fact is they didn't vote. And even if a non-voter would have voted for somebody if they were kidnapped and shoved into a voting booth, they still chose to vote for nobody. For some reason, though, Nobody is not allowed to win in elections.
So I got some data on United States presidential elections: voter turnout and results. I'm not concerned with whether a Democrat, a Whig, or a Republican won. Because of this, I decided to make these groups: all winners, all second-place finishers, and all others. Then most importantly, I added the voter non-turnout group.
Here's the graph you're probably used to seeing. Click the check box beneath the graph to reveal the graph I intended on making.
Toggle the check box on and off. You can treat the non-voter group (VAP) as saying, "I vote for none of the above candidates." This is, of course, treating a passive act as an active one, but it's still revealing because it shows that the non-voters outrank the voters for the winner in all elections after 1904. You read that correctly: every winner of a presidential election from 1908 onward has not received plurality of votes from those eligible to vote.
One might go so far as to say the president has not represented the people he serves for over a century.
As a final note: this data is just the popular vote and ignores the electoral vote, which is why the Loser beat the Winner in four elections (1876, 1888, 2000, and 2016) and still lost.