The O'Reilly Factor's online poll showed 92 percent of respondents believing George W. Bush will win reelection. John Hawkins of RightWingNews.com puts George W. Bush's reelection chances at seventy-five percent. I'm not convinced.

Today is election day. My prediction is that the election will be close. If someone were to put a gun to my head, I would predict that John Kerry will be elected the 44th President of the United States. The reasons leading me to this conclusion are passion, momentum, and history--three things John Kerry has going for him.

Passion "We are going to have an unprecedented turnout--the largest ever--and that is not good news for the GOP and President Bush," John Leboultillier writes in the most insightful analysis of the 2004 election I've read yet. Leboultillier points out that absentee ballot applications and early voting are at all-time highs, and that a considerably greater number of viewers tuned-in to the debates this year than four years ago. Voters content with the status quo don't flood polling places.

Momentum National polls are trending Kerry. The last twelve national polls featured on RealClearPolitics.com show Bush with an average lead of a little more than one point. The twelve previous polls gave Bush an average lead of about three points. However so slightly, the momentum has shifted to Kerry going into election day.

History Followers of elections know that undecided voters normally swing to the challenger in overwhelming numbers on election day. Followers of this election know that there aren't a whole lot of undecided voters. Still, Kerry benefits from this virtual certainty. Consider that in the last four elections featuring a sitting president, challengers gained an average of four points on election day from their election-eve polls while incumbents remained stagnant.

Should Bush eke out a victory, I will not be shocked. Should either candidate win in a landslide, I will be. Polls and pundits don't determine elections. Voters do. To quote Mills Lane, "Let's get it on."

I could see Kerry winning in a landslide but I can't see Bush doing that, therefore I think Kerry will win as well.

Let the blood-letting in the GOP begin. Maybe the Rockefeller Republicans and the neocons will get their come-uppance.

If Bush wins by a slim margin then Democrats will go batshit insane.

Dan...

I wasn't familiar with you until I picked up your latest book at the bookstore... almost bought your book... but, thought I'd check you out on the web, first. But, after seeing your comment implying Bush is a "turd sandwich" I'm not so sure.

In my opinion the key state in this election is Florida. Over the last 10 Presidential elections the winner has carried Florida 9 times. Even with Florida being a swing state Bush will win it by a very small margin and therefore win the presidency.

Is that a longing in your tone? A desire for a Bush loss (and therefore, a Kerry victory)? "That'll teach those silly neocons; I told you so!"

Sorry, the last post was by me; I meant to put 'Brian' at the top of the post.

I was confuesd for a second, I didn't know if DB was pulling something or if I had been posting in my sleep again.

I would say the longing is for the neocon day of reckoning but not for Kerry being president. The two are separable.

I actually want Bush to win, I really loathe Kerry b/c of his false Catholicism. My gut instinct tells me Kerry is going to win which I indicated by the mental test of which one I can see actually getting 300 electoral votes or more. I am just trying to be realistic or objective there. I can't see Bush pulling that off, if he did I would be quite surprised. I wouldn't be surprised if Kerry did as I really think there is real dissatisfaction with Bush that helps whoever is "not-Bush" on the ballot.

So the desire to see the neocon bloodletting is for the day after the election as a necessary retooling of the GOP. That way they can get on the right track for the future and make Kerry another one-term president.

According to Rush (so take this with a grain of salt), the polls are not as close as they seem for the following reason: pollsters have been assigning the 'undecided' voters to Kerry at a ratio of 1:9. They have been doing so, according to Rush, based on an outdated model that 'undecideds' vote overwhelmingly for the challenger (a view repeated on this blog). In the last three presidential elections involving an incumbent, however, the undecideds have split evenly. Thus, on a more accurate prediction model, the ratio should be 5:5, and consequently, Bush should be over 50% in every poll and leading Kerry by an average of two additional points. Can anyone spell "landslide".

Correction: make that at a ration of 9:1.

I would like to go on the record right now and declare John Kerry the winner and 44th President of the United States.

Brad, your figures are incorrect. Undecideds have indeed split for the challenger in recent presidential elections. I link to the data in my post (which is linked to within this post) on this election-day reality. My sense is that if Bush were the challenger and Kerry the incumbent, you'd be agreeing with me on this 100%.

Your sense is misguided. I am merely relating what I heard on the radio. Rush's argument can be found on his website:

http://www.rushlimbaugh.com/home/daily/site_110104/content/stack_a.guest.html

Rush cites data from the Univ. of Michigan that runs counter to your blog. Who am I to adjudicate? I am merely suggesting that your may not be the correct view. We'll know for sure tomorrow.

DAMN!!

Whether tonight's results conform or run counter to the historic trend is irrelevant to what is or isn't the historic trend. I looked at Rush's numbers, and while they relate to what I'm saying, his numbers are not on the same topic. To make it clear: polls undercount challengers.

Contrasting challenger performance on election day with challenger performance in the last polls prior to election day, you see a definite bump--about four points in the last four incumbent/challenger presidential contest. Incumbents generally remain at the same level as the election-eve polls. While this definite, clear historic occurence may or may not apply to what happens today, it's undeniable that it has applied to recent presidential contests--96, 92, 84, and 80.

My apologies; I misunderstood your point. In today's blog entry you write "...undecided voters normally swing to the challenger in overwhelming numbers on election day." I took this to mean that undecided voters could be expected to vote for Kerry in "overwhelming numbers," a fact challenged by the data Rush cites.

On October 18 you state it thusly: "Polls notoriously undercount challengers." You explain this phenomenon by connecting it to the voting trends of undecided voters: "This is mainly because undecided voters swing for the challenger on election day." Thus my confusion.

Common guys... look at the real facts.... Bush won the Halloween masks sales.

Sorry needed a bright spot here!

Although Rush and I aren't talking about the exact same thing Brad, what we are saying is in a sense in conflict. If polls undercount challengers, what constitutes the four or so percent boost that challengers receive on election day? Well, if it's not mainly undecided voters, then what is it?

I hear the pollsters are going nuts about everyone just using cell phones these days and so being unavailable for polling.

What with the "new" voters? How do they impact the election day bumps?