Now for the big enchilada. In 2004 I predicted all 538 electoral votes in each state, and predicted Bush's percentage in 45 states within 2% of the actual results (curse you, AL, MS, ND, RI, and VT!). So keep in mind that if you think I'm showing pro- or anti- McCain bias, my real incentive here is not to lose my bragging rights.
The state polling is interesting. The OxMethod weighted polling, then re-weighted by state population to make a national sample, shows a 50.9-43.7 Obama lead. This is akin to what the RCP average of national polling is showing right now.
The OxMethod polling presently shows Obama with a 353 to 185 electoral vote lead. What is interesting is that to get to electoral vote 269, McCain would only have to shift the country 5.7 points to the right. To get electoral vote 270 (and 271-274 as it happens), McCain would only have to shift the country 6 points to the right.
In other words, the tipping point for the electoral and popular votes do not match up. If McCain ends up losing the popular vote by around two points, he could still pull an electoral college upset. In other words, if Tarrance has it right, we could be looking at a situation that makes 2000 look like child's play. What makes this possible is a candidate who excites key Democratic demographics on the coasts and in the deep south, but not enough to flip that many electoral votes given some weaknesses he has in some the interior swing states. I don't think Tarrance probably has it right, but more on that later.
The number one factor in analyzing this election has been that no one -- NO ONE -- really knows what the electorate will look like. We've seen some turnout scenarios that look pretty implausible, such as Virginia flipping from slightly +R in 2006 to +10D in some SurveyUSA polling. But we can't really discount the possibility, given the historic nature of the Democrats' candidacy, and the external factors dragging down the Republican ticket. We hear anecdotal evidence going both ways on exit polling of early voters, and there's really no way to sort this out.
And we have a candidate on the Democratic side who has some history of underperforming the polls. Will it translate to the general election? We don't know yet. We know much about the Democratic turnout machine, just like in 2004. Will it perform (as it did in 2004), but find itself outmatched by an under-the-radar Republican effort (like 2004)? Will college students in Richmond and Philly get out of bed to brave a pretty crummy day? Will their increased numbers be swamped by the naturally increasing 65+ demographic? And how much are polls already "baking in" a massive, unmatched Obama turnout that may or may not occur.
We can't know the answer to these questions ahead of time, and it makes predicting things all the more difficult. The one bright side is that the last-minute swing in undecideds that was picked up in the national polling the last couple days can't be replicated in state polling. This is because the state polling generally does a better job of pushing undecideds, so there are fewer leaners to allocate.
Without further adieu, here's my breakdown, with analysis when helpful:
AL: 63-37 McCain (9, 0)
AK: 59-41 McCain (12, 0)
AZ: Undecideds here break against the well-known McCain for a 52-48 win. A major Republican headache in future years. (22, 0)
AR: 54-46 McCain (28, 0)
CA: 62-38 Obama (28, 55)
CO: This is an interesting state. As Geraghty has discussed, the Democrats' early voting edge has not been through the roof here -- it is improved to be sure, but not overwhelming. How much are pollsters already counting on such an edge for Obama? Call it a surprsingly narrow 52-48 win for Obama. (28, 64).
CT: 58-42 Obama (28, 71).
DE: 64-36 Biden (28, 74).
FL: I had a really hard time picking this state. The OxMethod has Obama at 48.85%, and McCain at 46.8%. But Bush overperformed the state by over five points in 2004, and McCain has the momentum. Obama leads in early voting, but so did Kerry. Of course, early voters probably make up a larger portion of the electorate, making it harder to overcome the deficit I pick an upset for McCain here, 50.5-49.5%. (55, 74)
GA: Closing quickly, but not quickly enough. A surprising squeeker here. Apparently populist conservatives don't like Wall Street Bailouts much. McCain, 52-48% (70, 74)
HI: 70-30 native son. (70, 78)
ID: Too bad Romney isn't on the ticket to push McCain to 80% here. 66-34 McCAin (74, 74).
IL: 61-39 Obama (74, 99)
IN: An Obama upset here isn't out of the question, but in the end the state is just too red for him to overcome. 52.5-47.5 McCain (85, 99)
IA: It's odd that both campaigns have been here of late, given that Obama is up 54-39 in the polling. 57-43 Obama (85, 106)
KS: McCain 59-41 (91, 106)
KY: Obama doesn't get too many undecideds here. 59-41 McCain (99, 106)
LA: A big AA turnout is already baked into the polls. 55-45 McCain (108, 106)
ME: 57-43 Obama (108, 110)
MD: 62-38 Obama (108, 120)
MA: Surprisingly poor performance for Obama, given the national mood. 61-39 Obama (108, 132)
MI: People blame McCain's slide on his abandoning the state, but this puts the cart before the horse. 56-44 Obama (108, 149)
MN: SUSA poll looks like an outlier, but hopefully it isn't. But if I were about hope, I guess I'd be voting for the other guy. 55-45 Obama (108, 159)
MS: AA turnout keeps this surprisingly close. 54-46 McCain (114, 159)
MO: Close one in the polls. This is a gut call, but I give McCain the tiniest of edges. Possible litigation here if other states end up close. 50.3-49.7 (125, 159)
MT: Some cash and organizers close the gap, but not enough. 53-47 McCain (128, 159)
NE: See MT. 58-42 McCain. Keep an eye on the second district (133, 159).
NV: Geraghty makes a fine case for a McCain win, but I just don't see it. If Jon Porter hangs on in NV-03, look for him to be the next junior Senator from Nevada. 52-48 Obama (133, 164).
NH: Well, you have a primary come-from-behind win for McCain, and a primary collapse for Obama. Don't see it repeating here. 54-46 Obama (133, 168)
NJ: 58-42 Obama (133, 183)
NM: Neighboring state once looked good for McCain. Not any more. Still, I think he gets some undecideds here. 54-46 Obama (133, 188)
NY: This will not end well. 68-32 Obama (133, 219)
NC: Another close one, but Obama seems to have a ceiling at 49%. Incidentally, early voting here shows the limits of early voting analysis. The first day showed a huge AA turnout, which tapered over time. This indicated that it might be more of a case of AA's excited to vote for Obama and turning out really early than something that will be sustained into regular voting. 51-49 McCain (148, 219). It's not really an upset, but it is going against the .3% edge Obama has in weighted polling (problem is that the trend has been against him, so earlier polling is probably artificially inflating his present standing).
ND: See MT. Also note that Bush's performance in these states is probably inflated by cultural affinity that McCain (and Bob Dole, who nearly lost SD in '96) don't share. 52, 48 McCain (151, 219)
OH: Recent polling trend favors McCain heavily here, and Obama is generally stuck below 50%. I really want to call this one McCain, but I'm going to call it 48-52 Obama (151, 219)
OK: My home state comes through for J Mac, 63-37 (158, 219).
OR: No longer a swing state. 57-43 Obama (158, 246)
PA: If you watched the Morning Call tracking poll cut Obama's lead in half in the last week, you'll understand why McCain is spending so much time there. And that poll uses D+10 for its model; it was D+5 in 2006. SUSA has Obama +9 and barely above 50% in a D+13 model). An upset isn't impossible, but I still have to say 53-47 Obama. (158, 267)
RI: 60-40 Obama (158, 271)
SC: 54-46 McCain (166, 271)
SD: 53-47 McCain (169, 271)
DC: Forgot about this. 90-10 Obama (169, 274)
TN: 58-42 McCain (180, 274)
TX: 55-45 McCain (214, 274)
UT: See ID. 62-38 McCain (219, 274)
VT: 66-34 Obama. (219, 277)
VA: The partisan breakdown of the polling gives me pause, but I still have to go with Obama here. 53-47. (219, 290)
WA: 57-43 Obama (219, 301)
WV: 55-45 McCain (224, 301)
WI: Same day voter fraud won't be the difference maker this year. 56-44 Obama (224, 311)
WY: 60-40 McCain (227, 311)
National total: Obama 53.4%, McCain 46.6%
All totals are 2PV