We can talk about impeaching Obama after we’ve put Cheney and Bush in jail.

I’m actually pretty determined to go back and find my old predictions. I’m not that good with predicting what individuals will do (although I did call the no pardons approach of Bush when almost no one believed that). Here are some of my old predictions:

I know that presidential campaigns are often insular, but Palin must be delusional if she thinks she has a chance in 2012. Two weeks after the election, the conventional wisdom will be that Palin cost McCain his chance at the presidency. The press will go back to swinging on McCain’s tire swing and Palin will go back to Alaska to be an unpopular governor. Even though she’ll still be popular with the Republican base, she’ll get waxed by Huckabee in the primaries in 2012. Palin will be lucky to finish out her term as governor.

And here:

I’m going to quote some predictions I made earlier. On Sept. 12th, I got into a bit of an argument with Ian Welsh at the mothership about Obama’s chances. I said this:

Obama will win in an electoral landslide (350+ electoral votes). The pollsters will spend a couple of months wondering why their likely voter models were so wrong, even though they saw the movement towards Obama in the final two weeks.

The financial crisis pushed things Obama’s way earlier than I expected, but I stand by the outcome prediction.

On Sept. 18, I made this prediction in a comment at fivethirtyeight.com:

The election won’t be that close. Obama will win and the only thing in doubt is the size of his victory. I’m putting this prediction in a comment so that I can point you all to it on Nov. 5.

Obama will win by at least 7 points more than his national poll lead two weeks before the election (unless he’s already ahead by 10 points by then). This will happen because the remaining undecideds will break his way and the +3% due to his ground game that the polls are missing. Undecideds will break his way because the economy sucks.

Best case for McCain: he polls 45% and third-party candidates get 5%.

Best case for Obama: he gets 57%, McCain 40% and third-party 3%.

Likely outcome: Obama 54%, McCain 43%, Others 3%.

Again, the financial crisis pushed people towards Obama earlier than I expected.

On Oct. 1, I sent my son an email with an optimistic electoral college map that showed Obama 376 – McCain 162. Obama carrying Kerry states plus Florida, Ohio, North Carolina, Virginia, Indiana, Missouri, Colorado, New Mexico, Nevada, and Nebraska CD 2. Today, I’m doubting he’ll pick up that 1 EV from Nebraska. On the other hand, if things really break his way, he picks up Georgia and North Dakota. Arizona would be just too much to ask for.

Now take a look at Sam Wang’s meta-analysis of state polls and realize that I made those first two predictions when things looked pretty much dead even in the polls and Obama was trending down. You can say I got lucky because the financial crash or McCain screwed up or whatever. Maybe you’re right. On the other hand, I’ll assert that it is possible to build a theory of how presidential elections work that allows one to differentiate between events that have transitory effects and those that move the polls in substantive ways. I called this election over on August 29th after McCain picked Palin. There was simply no way for McCain to win after that. It was pretty obvious that Palin would kill him in the states he needed to win.

The only state I missed in my prediction was Missouri and I always knew it was a tossup. I’m not convinced a recount there wouldn’t have put Obama over the top. I have incorporated a lot of post-election data into my predictions for 2010 and 2012. I think I understand much better today than I did back then about how things work. I’m willing to put my predictions up against any of the pros.

You think elections are determined by ideology, but that’s clearly not true. People develop their political identity in their late teens and early twenties. That political identity pretty much establishes their voting behavior for the rest of their life, barring life-changing events. Unlike 40 years ago, there are effectively no swing voters at the Presidential level. Simple demographics tell you that Obama will win reelection in 2012 with a larger margin than he had in 2008. Sure, it’s possible that some huge event could change that, but those things are really unlikely.

Mid-term Congressional elections are harder to predict, but the safest prediction is that the result won’t be more than a 2-3% swing in either direction. Given their incumbency advantage, that makes the Democrats a lock to maintain control of the House in 2010.