Bush wins Ohio! .. no, not yet.. yes... no - 2000 all over again?

No, it is not 2000.

Fox News and NBC are calling Ohio for Bush. With almost 90% of the precincts reporting, the margin is still the same... and it will stay the same.
Now, the other networks are not calling the state yet because the Kerry campaign announced that they were not going to recognize any result yet, until every vote, including provisional ballots, were counted. That actually creates a problem, because, by law, Ohio cannot count provisional ballots until eleven days after the election!

New Hampshire is not being called yet - that clearly represents a bad signal. I'm pretty sure that Kerry is going to win New Hampshire at the end, but I would haver considered as a good omen having that state called early.

Wow.. Louisiana is about to send their first Republican Senator ever. It was plain dumb from the Democrats not to unite under a single candidate when they saw that David Vittter was edging the 50% in the polls. I know that in Louisiana they usually conduct their races in this way, but, for almost two weeks it was becoming apparent that the Republicans could win in the first round, and yet the Democrats insisted in keeping divided.

It seems like the election is over for Kerry now, even if they are still trying to include all the provisional ballots. It doesn't add up. And Bush is getting ahead on the national popular vote with a comfortable three point margin. It's over because by losing Ohio, even if Kerry can win all of the other states (except Alaska of course), there would be a tie in the electoral college 269/269.

It is almost over

Some of the networks finally called Florida for Bush. Kerry is getting results in Florida way below expectations (I think that Mason-Dixon had Bush four points ahead - very close to what the results are looking right now, and living up to its fame as being the best pollster for that state, and for state polls in general).
Now it goes all to the 277-261 scenario for Kerry, but things are not looking promising in Ohio and New Mexico.
Around 85% of Ohio precincts have reported its results and Bush is still ahead by around 130 thousand votes - that, according to knowledgeable sources right beside me, may be a large enough difference to give the state to Bush, even with some results still coming from Columbus and other cities.
If Bush wins Ohio, the election is over, because in the best scenario for Kerry (winning the rest of the states) - the electoral college would be tied (269/269), throwing the election to the House of Representatives.

The Republican party is picking up Senate seats in Georgia, South and North Carolina. It seems like the Democrats are going to pick up Colorado. The Florida and South Dakota races are still too close to call.

Nothing resolved yet

New Hampshire is not called yet, only 5,000 votes differenece.
Kerry looks good in Pennsylvania but Florida is starting to look bad for him. Half of the precincts still to report in Palm Beach, one third in Miami-Dade and 13 percent in Broward. Kerry is behind 250,000 votes, if he wins is going to be very very very close.
Nothing yet in Michigan, Ohio, Wisconsin and Iowa.
No surprises yet on the Senate races - it seems that Bunning is going to win by less than one percent, but still win.
In Texas the races are very interesting. In Austin it seems that Mark Strama could unseat Jack Stick; the race between Baxter and Kelly White is tied! and Patrick Rose is looking as is going to be reelected.

First Results

I'm back!
We have the first results.. no surprises whatsoever. All the states that were supposed to go to each of the candidates has fallen that way. I'm slightly nervous about New Hampshire because it has not been called yet. New Hampshire is one of the closest states, but I'm looking at it as a signal for Kerry.
I'm pretty excited: the Kentucky Senate Seat is tied! 50-50%!! That was my upset prediction!
91% of the precincts counted .. and Bunning is ahead by some hundreds votes yet.

Final afternoon update

Soon we will start getting the results of the first states. Indiana and Kentucky are the first states that will report results, since polls just closed there. Bush is naturally going to win those two states, that are traditionally Republican.

This will be my last post in the next two hours (unfortunately), but I will resume later.
Ohio is going to be tight. The bets are still on Kerry, despite Drudge's warnings.
I would say that Kerry supporters should be cautiously optimistic, but the night is just starting.

Afternoon Update #3

Now it seems that my 311-227 alternate prediction is starting to look better. But, still way to go.
An aclaratory note as well. Some people asked me why I have the Democrats in red and the Republican in blue. The answer is that before 1996, most of the networks usually assigned those colors. Something happened in 1996 and now we call the Republican states as red (maybe because of red meat?) and the Democrat states blue (maybe because of blue blood?) - who knows.
I decided to stick to tradition (I've always used the colors I'm using today) and the website from where I get the maps, also uses these colors.

Afternoon Update #2

Wow.. the future markets are going crazy! Kerry up 15 points in 2 hours now!

Exit polls are not trustworthy when an election is very close, so we need to use caution here. There are still some hours for people to go voting, and we don't know the composition of those late voters.

It could be that Kerry is indeed winning by a more comfortable margin - in the 2.5-3.5% - but it is not over. Very good early news could discourage Kerry voters to go out in the final hours.

Afternoon Update

The first exit polls were released - Kerry is ahead. Of course, we need to be very cautious - the exact composition of the polled voters has not been released - only that the majority are women, which would give an expected advantage to Kerry http://drudgerepot.com
I actually feel much more comfortable with checking the financial markets (http://tradesports.com) - the price of Kerry contract has jumped 9 points in just one hour!

Kerry 252 - Bush 286 

Kerry 311 - Bush 227 

Kerry 277 - Bush 261 

Some hours before the election 

Who is going to win?

We start election day with no clear indication as to what the final result will be. As anybody who has being following the campaign season knows, the polls have been jumping all over the place and giving contradictory, odd and outright bizarre results. If you just follow the polls, then there's no way in which you can get an accurate prediction, so you need to include many more factors. Even doing that, it still doesn't give you a conclusive result.

But well, we need to have a prediction. Since last Friday I have been feeling a surge towards Kerry. I include here my prediction and try to give an explanation for them.

We basically enter election day with four states exactly tied on the polls: New Mexico, Iowa, Ohio and Florida, with 59 electoral votes among them. My prediction is that at the end of the election journey, Kerry is going to win New Mexico and Ohio and Bush will get Iowa and Florida. It well may be the other case around. I think that Kerry just slightly pulled away in Wisconsin, so I took that out of the final toss-up states.

I think there is a 70% chance of having a very very close election (less than 2% difference in the popular vote) and 30% chance of having a winner with a margin of 3-6%, which would give any of those a slightly more comfortable lead on the electoral college.

My prediction for the very close scenario is a Kerry victory of 277-261, and around 1.3% difference in the popular vote. If Kerry wins the popular vote with that 1.3% margin, then I think he will win Ohio and New Mexico (25 ev) and Bush will win Iowa and Florida (34 ev). If Kerry wins by 2-3.5% he will win all of those four states and their 59 ev, totaling 311 vs 227 for Bush. The same goes for Bush - if he can pull a victory in the popular vote between 1.5 and 2.5%, he will win those four states, giving him a total of 286 vs 252 for Kerry.

An extra point, that is, winning by 4.5%, would give Kerry a victory in Nevada and take him to 316ev - his total maximum. Two extra points to Bush (that means a victory of 4.5%) would give him the first 59 ev plus Wisconsin, Minnesota, Pennsylvania and New Hampshire - 45 extra votes to take him to a total of 331 vs 208 for Kerry.

Summarizing, my prediction is: Kerry 277 - Bush 261
With the following intervals:

2.5% victory for Kerry: 311-227
4.5% victory for Kerry: 316-222
2.5% victory for Bush: 252-286
4.5% victory for Bush: 208-331

You can check the maps here (above or below, don't know where are they going to appear) and the results of my model here: https://webspace.utexas.edu/cam949/www/Final%20Forecasts.pdf

I am also predicting the Senate and Governors results. I predict a pick up of one seat for the Republicans in the Senate.

Why I believe that Kerry has the advantage? Because of turnout. I think that we are going to see a considerable increase in turnout across the country. Another prediction: turnout will be at least 64% of the voting age population (against 54% in 2000). The range could be between 122 and 128 million voters (against 105 million four years ago).
I think pollsters have been having a lot of problems trying to measure all those first time voters, and what it has been widely talked about, those who only have cell phone lines instead of land lines, and who, by law, cannot be reached. Estimates range between 5 and 10% of the voters! Which is a a large enough interval to estimate anything with accuracy. A huge turnout will simply make the polls explode, and we could even get very surprising results. The most probable scenario is that both parties are energized enough and will get out to the polling booths in equal numbers.

If you analyze the market indicators, you can see a surge towards Kerry in the last 72 hours as well: http://tradesports.com

And finally, the Redskins lost their game against the Packers, and that signals a defeat for the incumbent party in the last 18 Presidential elections! - although, the Red Sox broke their curse this year as well.

This maps were created using the calculator tool at Dave Leip's Atlas of U.S. Presidential Elections (http://www.uselectionatlas.org), one of the best election resources on the web!