Jill Lawrence over at National Journal writes up the Harvard campaign postmortem by organizing it around four things that, she claims, "changed the course of the race and possibly even of history." Via Ezra, who tweeted that "You can never know for sure, but I doubt any of these things would have changed the outcome of 2012."
Yup. All four are worth thinking about.
One is that without his back issues, Rick Perry might have run a much better campaign and been nominated. I think that's probably a plausible story -- but it's hard to see him doing a lot better than Mitt Romney in the general election. Well, let me back up a bit.
Beyond the fundamentals of the campaign, which seems to call for a narrow Barack Obama victory, it seems possible that Mitt Romney was somewhat hurt by the unpopularity of the GOP. But he didn't himself have a reputation to overcome as an ideological extremist; the party did, but he was somewhat shielded from that.
So the question is whether a GOP candidate could have emerged with less of an ideological problem. Could Perry? Seems very unlikely; his campaign was entirely based on going to the right of Romney, Pawlenty, and any other mainstream candidate.
What about Romney himself? Another items basically suggests that -- the possibility that Romney could have taken a more moderate position on immigration. But in fact Romney's immigration position was, if not entirely forced, certainly a logical reaction to where the energy was within the party. On immigration, as with every other issue, there just wasn't room for Romney to moderate and still win the nomination.
The other two points, I think, are even less convincing. Lawrence cites the Ames Straw Poll's roll in knocking out Tim Pawlenty, but I'm pretty sure that gets the causation wrong; Pawlenty did poorly in Ames because his campaign had stalled overall. Yes, I'm sure Romney's campaign was worried about Pawlenty; who else would they be worried about in spring 2011? But they beat him, and while Ames may have provided a marker, it's unlikely it was really an important causal factor.
Then there's Bain. Yes, it was a vulnerability for Romney, and may have even hurt him a bit, but every candidate has vulnerabilities. It's hard to, in the abstract, guess at whether a particular attack hurt a candidate more or less than some other hypothetical attack on some other hypothetical candidate.
I really need to write a post at some point about how to think about election results and the fundamentals when fundamentals models diverge some, as they did in 2012. But overall, the consensus of the models certainly pointed to Obama winning, if not by as much as he did. My general sense is that the key factor (most of) the models don't account for had nothing to do with the individual candidates and instead were about the GOP itself -- both its reputation for extremism, and its reputation for botching the economy and Iraq. I don't have hard evidence for that, but it's consistent with all the evidence I've seen so far, and my guess is that we'll see it when more careful studies of the 2012 election show up in a while.