How did Donald Trump get elected president of the United States?   The simplest obvious answer to that question is: the electoral college. It needs to be emphasized over and over again that Trump didn’t win even a plurality of the popular vote, which went to Clinton by a margin that may reach 2 million votes when all the ballots are counted. Trump got about the same percentage of the vote that Romney got in his 2012 loss.

But of course, that answer is not sufficient. A habitually lying, malicious, hatemongering buffoon should not have been a serious contender, but he came close enough to win the electoral college. How?

Another simple answer is that history was against the Democrats. Historian Allan Lichtman had predicted Trump’s victory using a set of 13 key true/false questions, developed from historical observation, which has enabled him to correctly predict all the presidential election outcomes since 1984. Gimmicky as it may seem, Lichtman’s schema has merit. Still, it’s not really satisfying as an explanation for what happened this year; it leads us to view what this election had in common with past experience while underplaying the special circumstances in force today. We want to understand more than the answers to Lichtman’s 13 questions can suggest.

In any given election, personalities matter a lot, and that was certainly the case this time. Jonathan Alter is correct, I think, in saying that Trump didn’t actually win the election so much as Clinton lost it. Hillary Clinton is a brilliant woman with a superb presidential resume, but she was a very weak candidate, largely (but not entirely) for reasons that weren’t her fault. The fact is that years of right-wing propaganda since the early 1990s aimed at both Clintons, but especially at Hillary after she became the presumptive Democratic nominee for 2008 and again for 2016, were effective. The mainstream media, which were long addicted to reporting on Clinton pseudo-scandals, also contributed critically, most recently in playing up Clinton’s e-mail imbroglio and Clinton Foundation troubles far beyond their objective importance. Hillary entered the presidential race contending with a wide ly diffused, vague sense that she was somehow untrustworthy, shifty, or even corrupt. (In one of my worst ever predictions, I told a friend early in the 2000 decade that Hillary would never run for president, because she must surely understand that the widespread animus against her was an insuperable obstacle.) Her e-mail private server blunder was the icing on the cake for the right-wing propaganda mill, and Comey’s letter was the cherry on top.

Trump’s candidacy, on the other hand, was both strong and weak. Credit should be given where due: Trump is an extremely gifted demagogue, with a great ability to sense what his constituency wants and pander accordingly. His speeches effuse anger, disdain, and disgust, sentiments that resonate beautifully with the political sentiments of a significant portion of the electorate. And Trump is a highly talented showman, a talent that was the basis for his impressive skill at manipulating the media. Despite his phony complaints about media bias, the media actually played a big role in Trump’s success. Entertainer that he is, he was great for the media’s business, and he got tons of free air time practically from the moment he declared his candidacy.  The coverage almost invariably glossed over his mendacity and lack of substantive knowledge–the show was more important.

I had expected that Trump’s weaknesses—his obvious deficiencies in character, temperament and knowledge—would be fatally disabling, certain to alienate a large number of normally Republican voters. The Clinton camp were counting on this.  There were, indeed, significant defections among Republican political leaders and intellectuals, but ultimately, those defections didn’t amount to much in the polling booth. Trump won the support of 90% of self-identified Republicans, compared to Romney’s 93% in 2012. And Trump the outsider was clearly better positioned to exploit the current mood for change directed at an establishment that Hillary personified.

The election outcome has been widely interpreted as a revolt by a white working class that felt neglected by said establishment—a populist explosion of rage against the elites. There is much evidence to support that interpretation. Exit polls indicated that Trump won the non-college-educated white vote by a stunning 39-point margin (67-28%), outperforming Romney in 2012 (61-36%) and every other Republican presidential nominee since 1980. It wasn’t just among working class whites that Clinton fell short; she underperformed Obama’s 2012 advantages in just about every demographic category. She came in at 7 percentage points lower than Obama among black men, perhaps unsurprisingly, but even 3 points behind him among Latino men and 8 points worse among Latino women, despite Trump’s immigrant bashing. Her gains among women generally (one percentage point better than Obama) and college-educed whites (10 points) didn’t compensate for the other losses. So, while Clinton’s shortfall among working class whites was particularly marked, it wasn’t a complete outlier. But it seems to have been critical in the rustbelt of declining manufacturing areas: it apparently turned around three states—Pennsylvania, Michigan and Wisconsin—that Democrats normally count on winning in presidential elections. With those three states in her column, Clinton would have won.   Working class whites were also clearly important in flipping two other states—Ohio and Iowa—that Obama carried twice.

So, yes, a revolt by the white working class was a decisive contributor to this election result. How to explain that revolt, and how Democrats need to respond to it, will be the subjects of another post.