After the 2016 election, once everyone recovered from the shock, the analyses of what happened started to gel into a conventional wisdom that argued Donald Trump won because a bunch of non-college-educated white people felt "economic anxiety." So we got thousands of stories and features from reporters sent out to rural Pennsylvania diners and Iowa church socials to figure out what Trump voters truly want. But the fact was that it was an extremely close Electoral College victory that could have gone either way with just a handful of votes in a couple of swing states. The main data guru at the time, Nate Silver, did a post-election analysis which showed that whenever there was an event, such as Hillary Clinton collapsing briefly at a 9/11 event or the Washington Post reporting of Donald Trump's gross commentary on the "Access Hollywood" tape, there would be a slight drop in the polls for the affected candidate but they would rebound to the usual stasis within a couple of weeks.
Trump was still struggling to recover from the "Access Hollywood" scandal at the end of October of 2016 and Clinton was ahead in the aggregated polling by about six points. Then FBI Director James Comey sent a letter to Congress announcing that the agency was following up on the Clinton email investigation, and the media once again went wild with the story that had captivated them for months. Clinton's polls immediately dropped and never had a chance to recover because Election Day was just too soon. The rest is history.
