Right up until election day, most people thought Hillary Clinton would win the 2016 election. She didn't, and many people have many different theories about what went wrong. Which one you ascribe to probably does a fair bit of work in explaining who you back in 2020 and why. And the fact that we don't really know which answer is the right one, and feel an urge to cover all our bases "just to be sure", probably contributes to the Johnny Unbeatable problem -- the explanations aren't all mutually compatible, and so "learning their lessons" will pull us in contradictory directions.
Theory #1: It was anti-Clinton mania: Claim: Hillary Clinton is uniquely reviled for idiosyncratic and effectively non-ideological reasons. Lesson: don't nominate Hillary Clinton. Simple.
Theory #2: It was misogyny: It's not just Hillary Clinton. Any female candidate is going to whip up a toxic brew of wounded masculinity and machismo. People nervous about nominating another woman often believe this.
Theory #3: Clinton was too elitist: Generally a favorite of not-Democrats, though endorsed by some "moderates" too. The idea here is that Hillary Clinton represents a far-left agenda and the Democrats need to return to the good old days when they were the party of ... Bill Clinton? JFK? FDR? Unclear. What is clear is that Democrats need to show more respect for rural heartland voters, respect people who oppose gay rights, respect people who want to deport immigrant children, and above all respect people who want to ban abortion. Oh, and under no circumstances should a Democrat call anyone "deplorable" -- except maybe Ilhan Omar. These folks believe Joe Biden would have won in a cakewalk
Theory #4: Clinton was too centrist: Hillary Clinton didn't offer a true progressive alternative to conservatism, to capitalism, to corporatism, to technocracy, or to neoliberalism. Hence she didn't inspire voters thirsting for an alternative. Donald Trump at least purported to speak to voters who believed the system had failed him, and a Democrat who basically was seen as a "the status quo is okay" figure would not do well. A bold and uncompromising progressive vision that unabashedly targets the systemic forces immiserating us all, by contrast, can speak to the millions of Americans who feel dejected, disempowered, and ready for a change. Favored diagnosis of the "Bernie would have won!" crowd.
Theory #5: Clinton abandoned the Midwest: "She didn't even visit Wisconsin!" The blue wall cracked, Democrats lost Michigan, Pennsylvania and Wisconsin. So to win -- rebuild that blue wall, by nominating a candidate most likely to be extremely popular among rust belt swing voters in the Midwest. This could point you anywhere from Joe Biden to Tim Ryan to Amy Klobuchar to "recruit Sherrod Brown".
Theory #6: Clinton didn't excite the base: Democrats took for granted the constituencies that are most responsible for powering them to victory and so left a ton of votes on the table. Rather than chase the elusive (and probably white) swing voter, the strategy here is to goose turnout among voters (especially women) of color. It's not the same as Theory #4 because -- despite what Jacobin Mag would have you believe -- it is not necessarily the case that people of color are inspired by fire-eating leftism of the Jacobin Mag bent. Often believed by those who think Democrats must nominate a woman and/or person of color; and often also paired with a belief that Democrats should abandon the rust belt and focus more on winning emergent sun belt swing states like Arizona, North Carolina, or (dare to dream) Georgia and Texas.
Theory #7: PC backlash: The Democratic Party allowed themselves to be taken over by a radical identity politics fringe who alienated regular Americans with all this talk of "intersectionality" and "microaggressions" and "trigger warnings". Trump's victory was a result of a PC backlash prompted by White men who just was sick and tired of being called racist all the time, and lashed out by ... endorsing a really racist candidate. The interesting thing about this diagnosis is that its political description basically boils down to "White Americans are hella racist" while its political prescription is usually "don't ever say White Americans are racist." Facts sure care about those feelings.
Theory #8: Voter suppression: Lower turnout among the base wasn't (just) because of less enthusiasm for Clinton. It was also a function of a sustained conservative campaign to obstruct poor and minority communities from voting. Proponents of this view aren't necessarily tied to a particular candidate as they are to urging Democrats to take voter access seriously as a top priority -- both in terms of legislating and in terms of activism. Unfortunately, the Senate won't pass any legislation, the courts are basically endorsers of the voter suppression project, and it's not like Republicans are going to be less invested in voter suppression the next time around, so this can rapidly turn fatalistic.
Theory #9: Conspiracy theories: Russian interference, social media, fake news. Voters were buffeted by a frenzy of misleading or outright false information. Not only were some taken in by outlandish nonsense (QAnon, Pizzagate, Soros conspiracies), even those who weren't directly affected often suffered a general decline in trust for political institutions and the reliability of authority -- something that ended up redounding to Trump's benefit. And the Republicans exploited media timidity and its reflexive instinct to tell "both sides" to put naked falsehoods on par with actual truth. To a large extent, people in this camp think the media has to accept in a much more decisive manner its obligation call lies lies -- even if it makes "certain" elected officials mad.
Theory #10: The Democratic Party is corrupt/incompetent/in disarray: Democrats don't really want to win, or they don't want to win if it means displeasing their true corporate masters. The party is beholden to an out-of-touch consultant class and big money donors locking them into unpopular positions that predictably lose elections. Only by seizing control of the party apparatus and smashing the old guard can Democrats actually run races that will actually inspire people. The more militant version of Theory #4.
Theory #11: Nothing went wrong -- it was a fluke: During the entire 2016 cycle, the polls oscillated between a high of "Hillary Clinton landslide" and a low of "statistical dead heat". The argument here is that all that happened in 2016 is that election day happened to have the tremendous bad luck of occurring at one of the nadirs, leading to what was effectively a statistical dead heat and a coin-flip Donald Trump win. The lesson here is that there is n: o lesson: if you have a campaign strategy that gives you a win 75% of the time, that's a pretty good strategy even though (as Nate Silver reminds us) statistically you should lose a quarter of the time. All the efforts to "explain" the outcome basically a function of statistical illiteracy. To the extent Democrats should be on alert for anything, it's in letting the "we have to change something" impulse to tinker lead to making things worse.