Notice that I haven’t even gotten to the e-mails — a story that broke in March of last year, the point at which it was still, if barely, possible for someone to sit down and say “for the good of the party, I’m afraid you need to develop a serious health issue that will require you to drop out of the race.” Instead, the left half of the political spectrum solemnly agreed with themselves that this was a non-story, a three-day wonder, nothing for us to worry about. I am sympathetic to this decision only because by that point there was no one else, the entire machine of Democratic politics having been turned into a purpose-built vehicle for the Clinton candidacy. But why did the Democratic power brokers settle on Clinton back then? She could easily have ended up facing a much more viable opponent, like Scott Walker or Marco Rubio or John Kasich.

How can we explain this? For one thing, I think Clinton’s candidacy — like Trump’s candidacy, in its own, very different way — points to the fatal weakness of the political parties. Decades of “good government” reforms have systematically stripped the power that parties once had: to control money, to control committee assignments, to control how much pork politicians get to brag about to the voters back home. What’s left is a hollow shell that cannot effectively respond to either grassroots insurgencies or to outsize figures who effectively turn the party apparatus to their own ends. If you think, as I do, that parties play a vital role in organizing political action toward coherent goals and long-term accountability, that’s something that should worry you.

My other working theory is simply that the left committed the cardinal sin known as “reading your own press releases.” The left loved the “demographics is destiny” arguments that seemed to promise them a glorious future of uninterrupted rule. I cannot count the number of liberals who solemnly assured me that there was not going to be a Republican president in 2016, or any other year in the foreseeable future, because demographics.