I know people think that my life is all Stata, marmots, and Michael's Frozen Custard, but in addition to this, I am also currently following American electoral politics obsessively. This includes several attempts to figure out and specify, to my personal-politico-epistemological satisfaction, Kerry Victory Scenarios.
I recognize that this post is me doing public political science, for which I am unlicensed, rather than public sociology. The advantage of my analysis, as opposed to more expert opinions you may read, is that it is done with JFW's promise of being conducted without spin or romance, and purged of illusions as far as I am able. And it's not like I'm entirely inexpert on matters of numbers and surveys and social trends.
In the official JFW reckoning, the Kerry Victory Scenarios boil down to the question of how Kerry can get 38 electoral votes from the following six states: Florida (27 EV), Pennsylvania (21), Ohio (20), Missouri (11), Iowa (7), and New Hampshire (4). Really, I think, despite various depictions of the "battleground states" as numbering into the teens, those really stretch into the periphery of the battle. Meaning, for example, that I don't count Wisconsin as a core battleground state, not because Kerry can't lose Wisconsin, but, if he loses Wisconsin, I don't see him winning 48 (38+WI's 10) EV from the six states above either.* Likewise, Bush could lose Virginia, conceivably, but he's not going to lose Virginia and not lose 25 (38-VAs 13) EV from the six states above.
Anyway, as I see it, there are basically seven Kerry Victory Scenarios, in order of my estimate of their plausibility. Italics indicate the state(s) that answer the key question to the formulation of any Kerry Victory Scenario, which is what state Kerry is going to win that Bush won (or, as some might say in one notable case, "won") last time.
1. Florida + Pennsylvania (48)
2. Florida + Iowa + New Hampshire (38)
3. Pennsylvania + Ohio (41)
4. Pennsylvania + Missouri + Iowa (39)
5. Florida + Ohio (47)
6. Florida + Missouri (38)
Even more borderline implausible:
7. Ohio + Missouri + Iowa (38)
So this is it. If I am correct: Kerry does not have to win Florida. He does not have to win Pennsylvania. He does have to win one or the other. He does not have to win, as some are saying, two-out-of-three of Florida, Pennsylvania, and Ohio. If he does win two of these three, he does win.
Tune back on November 3 to see if I know what I am talking about. My prediction being that, if Kerry wins, one of the these scenarios happened, and, if he doesn't, none of them did.
* Update, ten minutes later: Okay, okay, I guess I could also see Florida + Pennsylvania - Wisconsin or Pennsylvania + Ohio + Iowa - Wisconsin, so that makes Wisconsin a semi-core battleground state and increases the list of scenarios to nine. If you wanted an even ten, I think the most plausible scenario I've ruled as implausible is: Florida + Iowa + West Virginia (39; which is not to say that Kerry winning WV is implausible, but winning WV while losing PA [covered under scenario #1] and NH [#2]?)