> Rather than asking how Trump won, we should be asking how Trump was close enough to win.
... Here, then, is Bartels’s point: if you’d been given a printout of voter data from the past few elections and been asked to identify which campaign was the bizarre one, the one that would rock American politics and lead to book after book trying to explain the outcome, would you be able to do so? The results in 2016 mostly looked like 2012 and 2008 and 2004, even though the winning candidate is one of the most bizarre Йgures ever to crash into American politics.