In Part III, I essentially left off with the question of where the responsibility should be for electoral outcomes-- the voters or the candidates? Should an election be a game between candidates where voters respond mechanistically to candidates, or should it be an assessment of candidates where their actions are irrelevant, and the electorate judges them, disregarding the trivialities? (Hint: that's kind of a false dichotomy...)
My first book examined an election as a hiring decision (or a contract renewal decision). Whom does the electorate hire for an open seat, or do they fire the incumbent? Such decisions should be made on the basis of job performance, or objective predictions of potential performance. Fancy stationary for one's resume? If that affects the process, you've got problems. If one applicant, or an existing employee can dissuade other applicants from seeking jobs at a firm, you've got problems. Anything that pulls the decision away from actual job performance is a problem.
"Clinton didn't adopt the right platform" would be a strategic critique with normative implications acceptable within the bounds of democratic theory because it relates to job performance. "Clinton didn't go to Wisconsin?" Nope. She used the wrong stationary for her resume.
This really is that initial question here. When people ask whether Clinton is to blame for losing 2016, what they are really getting at is as follows. There are three actors/sets-of-actors. Clinton, Trump, and the electorate. Normally, we try to avoid treating a group of actors as a unitary actor, and I will try to avoid doing that here, but this matters. The "Clinton didn't go to Wisconsin" argument removes the electorate from the calculation in a way that indicates a failure of process. It absolves them of any responsibility for any decision by pretending they don't even have a part to play in the process. The election was just about Clinton's decisions, or possibly her Risk game with Trump. The idea of voters evaluating the candidates' resumes has no part in that conception, and that's the central problem with the "Clinton didn't go to Wisconsin" argument.
Now, that said, consider the pure spatial model that I have described. Suppose that all voters can be placed along a left-right continuum, and every voter votes for the candidate closest to him. The candidates are forced to converge to the location of the median voter, which has some nice mathematical properties. A candidate's failure to do so is a strategic failure that will result in loss. And the voters here are purely mechanistic. Candidates are strategic, voters are mechanistic, and we can be kind of comfortable with this system. Why? Because the nature of the mechanism is fundamentally about job performance rather than stupid trivialities like "Clinton didn't go to Wisconsin!"
I keep coming back to this because, yes, the electorate can be mechanistic, and yes, it can be stupid, but if the electorate were mechanistic in that particular way, we should be even more horrified. The idea that the election is based on Locke, Sawyer and Hurley playing Risk before... well, a bunch of bad shit happens is a conception of the electoral process in which there is no role for anything that can conceivably lead to good job performance.
And hey, did anyone notice that Trump doesn't know what the fuck he's doing? So, even if the election did turn on Clinton not going to Wisconsin, that kind of makes my point about how a hiring decision made on that basis leads to an employee who can't do his fucking job because he's so fucking stupid that he makes satire impossible.
Next, does any of this "she didn't go to Wisconsin" shit really matter? Probably not, but I'll get to that next.