Detailed Election Predictions

This election cycle is too complicated to simply make a single prediction, so I’ll start by laying out the three possible scenarios for tomorrow.

Scenario #1 (Pessimistic): Obama wins the national popular vote and over 300 electorsScenario #2 (Realistic): Obama wins the national popular vote very narrowly but the winner of the electoral college comes down to the outcome of three states Scenario #3 (Optimistic): McCain wins the national popular vote and the electoral college.

The first scenario is what will happen if the assumptions underlying the media’s polling this year are correct. This is, in my opinion, the least likely of the three scenarios. In 2006, the best year for Democrats in recent history, Democrats had a +3 advantage in turnout. The polls this year are assuming the Democrats will be advantaged anywhere from +8 to +16. If this is true then Obama will win easily and it will likely sound the death knell of the Republican Party. However, we know that the polls consistently overestimate Democratic support (partly because of bias and partly because of the “Shy Tory” factor: Republicans are less willing to talk to pollsters). We also don’t have any reason to believe that the Democrats will have this large of an advantage.

Scenarios 2 and 3 take account of the fact that Democrats are overpolled and that remains a significant percentage of undecided voters who are breaking for McCain. The last few days of the campaign have not been great for Obama (coal comments) but it is likely that the momentum for McCain was slowed by the passing of Obama’s grandmother today. How much the momentum slowed could determine this election.

Scenario 2, which I view as the most likely, assumes that Democrats improve their advantage to around +6 (which would still be an unprecedented advantage). Scenario 3 recognizes that 2006 was a horrible year for Republicans and that 2008 is bad but it is not worse, so we should expect Democrats to have a +3 advantage or less in turnout. Additionally, these two scenarios recognize that a substantial number of Clinton supporters and Reagan Democrats do not like Obama and will be voting for McCain (meaning he will win a larger share of Democrats than usual). This appears to be supported by exit poll results from early voting.

With all that said, on to the states:

Safe McCain States (McCain will win in all three scenarios): 174 EVIndiana, Kentucky, Georgia, South Carolina, West Virginia, Tennessee, Alabama, Mississippi, Arkansas, Louisiana, Texas, Oklahoma, Kansas, Nebraska, South Dakota, North Dakota, Wyoming, Montana, Idaho, Utah, Arizona, Alaska

Likely McCain States (McCain may lose under scenario #1): 73 EVNorth Carolina, Missouri, Florida, Ohio*FL and OH are likely to be closer in the final outcome, but they are states in which Kerry was doing better in the polls in 2004 than Obama is now and still went Republican easily. They are essentially must-wins for the latter two scenarios to be relevant, but I am confident they will again go Republican.

So, if scenario #2 or 3 is correct McCain is at 247 electors, 23 short of victory. Now things get interesting:

Lean McCain (McCain will probably win under scenario #2, will win under scenario #3): 5 EVNevada

Possible McCain (McCain could win under scenario #2, will probably win under scenario #3): 16 EVNew Hampshire, Iowa, New Mexico

This brings us to the three key states:

Key States: 43 EVPennsylvania (21), Virginia (13), Colorado (9)Under scenario two it is possible, but unlikely, for McCain to win if he just carries Virginia or Colorado. The likely path to victory, however, means McCain wins either Pennsylvania or Virginia AND Colorado. Either combination would put him on top.

This is the situation we are facing. With that, my prediction is that scenario #2 is most likely. McCain therefore carries all the safe states and the leaners. He also wins Nevada and Iowa. I think it is possible for McCain to win all three of the key states but I’m going to predict that the election will be decided by Pennsylvania which will go narrowly for McCain (I think Virigina will also vote for McCain but this simply pads his margin).

McCain’s Predicted Electoral Vote Count: 293

I’ll conclude with some possible surprises to look for: Maine 2nd Congressional District and Minnesota (a Survey USA poll put McCain within the MOE today). If McCain wins either of these my predictions will be happily wrong and we will likely be looking at scenario #3.