Recently in Election 2004 Category

Taggert Brooks posted this before the election.

I'm inclined to agree. Polls probably do influence voters. It is a lot like the Heisenberg Uncertainty Principle.

Let's apply this to exit polls. We'll probably never know, but I'd like to know whether Republicans came out later in the day after they heard early exit polls saying that Kerry appeared to be leading or that it was going to be really close. For more on this story, read here. Here is part of the article:

That said, Harris believes that the media and the campaigns misuse exit polls.

"It got on Drudge (Report), it got to the campaigns," Harris said of Tuesday's early exit polls. "You looked at President Bush (Tuesday) morning and he looked sick. Exit polls are a wonderful advantage in providing context and texture. But for trying to make a call, we've learned in two elections in a row that they lead to confusion."

Bill Schneider, a CNN analyst and former Harvard University political science professor, agreed.

"The lesson here is put not your faith in exit polls ... particularly if the exit poll is close," he said. "Exit polls are designed for analysis. ... They are not very good, nor really is any poll very good, for being an absolute, exact prediction."

But Karyn Barker, a fellow at the American Enterprise Institute, believes the exit polls are getting a bad rap. It's not the polls, she says, it's how the people use them.

"Anyone who follows exit polls should not use them as a prediction of the turnout of certain groups," she said. "To assume that these numbers were spot-on was foolish. It seems some people, including me, wildly misinterpreted what they meant."

Zogby said Tuesday's exit-poll problem would probably send pollsters back to the drawing board to figure out whether they need to apply different formulas, models and weights to their surveys.

"There was a time you could go to the bank with the early exit polls," Zogby said. "Now we have a problem."

I think they are missing the point as to what the real problem is.

The election has been decided. There's still more to talk about, and I might post links to interesting things I read. Voter turnout is going to be a hot topic. The errors of the of exit polls (and other polls) are going to be hot topics. And of course we can always talk about the role of the new media in politics. That should keep us occupied for 4 years.

Lots of economic news in the wings. More posts on that in the future.

The Fed meets on Tuesday Wednesday. To be continued...

The title of this CNN article should be "Stock Market Rallies on News of a Clear Winner." It would have gained 100 points for Kerry too. Only a recount would have sent it down.

The Iowa Electronic Market, which I have referenced often on this blog, did one thing very well. It consistently predicted Bush receiving the majority of the two party vote. Polling data had some problems this time around, that is for sure. Alan Bjerga mentions the polls among other things in his post-election reflection. But you never really heard much about nationwide polls this time. The polls were all focused on the battleground states. Maybe next time the IEM should do a battleground poll.

The winner-take-all market reflected the uncertainty in that prediction. Even though the vote share was dead-on, the winner-take-all made a big swing back and forth at the end, indicating Kerry before the polls closed, but then very quickly--and I mean you could watch it changing before your eyes--switching to Bush as the returns came in.

All in all, not bad for the IEM. I think it worked a little better than in 2000. In fact, this sort of result (51-49 or 52-48) seems to be where it is a really useful tool.

If you listen really carefully, you can sometimes hear the NBC chimes (G-E-C) in the background before a state is called for one of the candidates. For the history of the NBC chimes, go here. Besides being a symbol of the network, the chimes are an internal signal in the newsroom at NBC of a news bulletin.

Dan Rather is throwing out homespun sayings at a rate of about two per minute. "Just because a chicken has wings doesn't mean that chicken can fly," for example. (I may not have gotten the quote exactly, but you get the idea.) I have no idea what he meant.

I've really enjoyed NBC and MSNBC though. The analysts have personalities, and the panels have been pretty balanced--those are two things I like in election coverage. They put together a nice production overall.

I'm about to call it quits, but I have been studying the numbers for the last hour or two. It is hard to imagine Ohio switching to Kerry unless there is some sort of fraud allegation that hasn't yet been made. New Mexico and Nevada are leaning toward Bush. We'll probably wake up to see them in his column.

Now then, if NV and NM go to Bush and Ohio somehow goes to Kerry, then Bush would have to pick up Wisconsin to force a tie. Iowa (which has suspended their counting for the night, but Bush leads with Republican areas yet to be counted) would not be enough.

Wisconsin plus Ohio would be enough for Kerry at this point, and it looks like he'll get Wisconsin. That's why there will probably be some kind of challenge in Ohio, or at least the exploration of his options. For Kerry, it's that important.

Ohio.

Unless there turn out to be major accusations of wrongdoing in Ohio (and no one has reported anything like this yet), this one is over.

By my estimation, the final result will look a lot like it did in 2000. NH will turn blue while NM and IA will turn red. (Though technically all three of those are too close to call--my predictions.)

I love this stuff no matter who wins. I've been glued to the TV to see Reagan, Bush, Clinton, and Bush hit the magic 270. It never gets old. It never will. It is just fascinating to watch.

I'm going to watch until Tom Brokaw signs off. This will be the last election that he covers as an anchor. Get some sleep, Tom! Thanks for the memories.

Hilary Rosen on MSNBC says that Bush shouldn't declare victory until the votes are counted in Ohio.

Agreed.

I had forgotten that I made this trade back in January. I bought 15 shares of Bush vs. Kerry at between 16 and 17 cents each. They will pay off at about 52 cents.

That'll buy my lunch tomorrow!

But hey, it's fun.

Things seem to be breaking in favor of Bush. It will be interesting to see how close the big states (OH and FL) are to see if there will be any challenges. It will depend on the provisional ballots, absentee ballots, and so on. I'm watching NBC and MSNBC mostly tonight. Their analysis of Ohio has been absolutely fascinating. This will keep people talking for a long time.

The real story here is that the exit polls all night have been wrong. Theories abound.

Even so, I don't think Kerry will concede tonight unless Bush absolutely runs away with Ohio, and maybe not even then.

Bush and Cheney might address the nation tonight (this morning), this prediction based off of seeing the VP motorcade heading to the White House on one of the networks as I flipped around.

I'm ready to call it, but it is not exactly the way I expected to.

Tim Russert just said that is going to be very hard to put together a winning coalition of states without Ohio and Florida.

Indeed.

The title is the message.

Too early for me, but they are doing it.

Challenges are likely.

I like Tim Russert's new technology (notebook computer rather than whiteboard). When you listen to him run the numbers, he sounds like he is going through some internal struggle not to call this for Bush. I guess that's because he all but has Ohio in the Bush column. If that turns out to be true, Kerry would (almost) have to run the table. But I'm not convinced.

People are still voting in Ohio! Anyone in line when polls closed will vote, at least by provisional ballot. Here's to participation!

MSNBC called Florida for Bush

MSNBC reports from their decision desk that they may not call Florida tonight because of the absentee ballots.

Here is the current quote from the IEM website.

Quotes current as of 22:45:04 CST, Tuesday, November 02, 2004.
Symbol Bid Ask Last Low High Average
DEM04_G52 0.007 0.009 0.007 0.006 0.950 0.211
DEM04_L52 0.111 0.198 0.112 0.110 0.500 0.354
REP04_L52 0.671 0.693 0.693 0.000 0.700 0.277
REP04_G52 0.161 0.187 0.188 0.033 0.220 0.107

This is from the winner-take-all market. The most likely outcome according to the market is Bush with less than 52%. Seems a little high. Time will tell.

Florida is 97% in, but there are a lot of absentee ballots out there. Dee Dee Meyers on MSNBC thought there were a million of them. So it's clearly not over, even in Florida. I don't think I'm saying anything controversial here when I say that had it not been for 2000, they would have called Florida by now. But obviously no one wants to go through that again.

I'll say it again. Ohio is a must win for Kerry. But that one is far from over. I don't think we'll know the answer tonight. Lots of provisional ballots to count in Ohio.

Judy Woodruff on CNN just reported an unnamed Kerry source is pessimistic about Florida.

Nothing. Absolutely nothing firm to report.

Looks like Florida is tipping toward Bush (as I was expecting), but Ohio is really too close for anyone to even think about calling. With apologies to Tim Russert...

Ohio
Ohio
Ohio

And maybe Iowa, Minnesota, and Wisconsin. Very interesting indeed. The total on MSNBC is currently 193 to 112 in favor of Bush. Oops.. check that... news coming....

PA for Kerry
AZ for Bush

203-133 for Bush as of now.

And still no surprises.

Blogging into the night...

Dixville Notch voted. Bush 19-Kerry 7.

The rest of us do it in the morning.

The story goes like this: if the Washington Redskins win their last home game before the presidential election, the incumbent wins, and if they lose, the incumbent loses. It has correctly "predicted" the outcome since 1936.

It's amusing, and meaningless, of course. The two events have nothing to do with one another. But it does give us an opportunity for a little statistical fun.

Suppose the probability of the two events happening together (Redskins win/incumbent wins or Redskins lose/incumbent loses) is 1/2. It's not exactly, but it's a reasonable approximation for starters. The probability of it happening twice in a row is 1/4. The probability of it happening n times in a row is (1/2)^n. There have been 17 elections since 1936, so (1/2)^17=0.0000076. Impressive indeed!

Or maybe not. Open challenge (I might try to analyze this if I have time): argue that the success of this indicator of elections is not as impressive as the above paragraph suggests. I have the beginning of an argument started, but it's already way too long to put here.

It's kind of like the observation that an old NFL or NFC team winning the Super Bowl tends to coincide with a rising stock market that year. Historically, the NFL or NFC has won more Super Bowls (they have had better teams) and the stock market is up more years than it is down (for reasons that have nothing to do with football). Note that this predictor seems to work well only because the Pittsburgh Steelers (4 Super Bowl wins) are an old NFL team and correctly predicted bull markets! If the Steelers were counted with the AFC (their current conference), the predictor would not do much better than pure chance.

And that's the beauty of looking for these patterns. People just search around until they get something that works! That's why I am always so skeptical of stuff like this. Big deal! It's amusing--nothing more.

Bottom line: The world is full of improbable events. Improbable events are literally happening all the time. Each one is so unlikely, but there are so many opportunities for unlikely things to happen. Sometimes a coincidence is just a coincidence!

I have an academic paper on the subject of conditional probabilities, so this is more than just a passing interest. I might write more on this after the election--no matter who wins.

As I write, it is 8 1/2 hours to midnight in the east. 8 1/2 hours until the voting starts (and ends) in Dixville Notch, NH. The tiny town has a tradition of voting first in the nation. Some 20 hours later, the real fun begins.

I don't have a real prediction in the sense of saying that I think one candidate is sure to win. I do think that the popular vote will be decided by less than 2 percent in either direction and that a tie in the electoral vote is more possible this year than at any time in our past. I would put the probability of an electoral vote tie at about 20% right now. What, you ask, yields a tie? Among other things, Kerry takes PA, OR, WA, HI, OH, MI, ME, NH, and IA while Bush takes, CO, AZ, NM, FL, and WV. Notably missing from this equation are MN and WI. They would need to split those two (10 votes each) to end in a tie.

I don't think Bush can win without Florida, and with FL and Ohio he's unbeatable. Kerry can't with without PA, but I also don't see how he can lose his wife's home state. Iowa is also crucial for Kerry. Only 7 votes, but I think that the vote in Iowa will representative of what is happening in other battleground states. If Bush wins Iowa, he has a landslide. If Kerry wins Iowa, it's not over. Likewise, another small benchmark state will be WV. If Kerry can win there, he'll have the landslide. If Bush takes it, he's still in the game. Look for these kind of indicators tomorrow night to gauge where things are going.

Stephen Karlson quotes a Journal Sentinel article reminding us why the electoral college is a good thing. Their main point is that if the presidency were decided by popular vote rather than the electoral vote, there would be a major shift in the way campaigns are conducted. Big markets like California and New York would be the only focus since advertising dollars would hit more people than in, say, northern Iowa.

The real genius of the electoral college is that the vote reflects things about this country that would not come out of direct popular voting. It gives more weight to areas of the country that are changing. Florida is growing and becoming more diverse, making it a place worthy of attention by the candidates. Minnesota used to be a liberal stronghold. Now it's in play, and Bush visits almost weekly. Ohio, home of Taft and McKinley, is in play; Kerry has a chance there. As I mentioned earlier, Nevada is hard to read because of the way it is changing.

It's not perfect, but it is quite remarkable in the way that it puts the focus precisely on the areas of the country where divergent views seem to be concentrated. I think that this is necessary for the health of the republic if we are to remain the United STATES of America.

It's close here too.

And that goes for election prediction websites too. I'm looking at this as objectively as I can. Look at the contrast in the maps at electoral-vote.com and MSNBC. On Monday morning, electoral-vote.com had Hawaii in the Bush column. I don't think so. All I can figure is that this is just an example of how bad some polls can be. The same site has New Mexico and Ohio for Bush, albeit weakly. By the same token, I find it hard to accept New Hampshire being a tossup state as it is on MSNBC's website.

As I see it today, New Mexico is a tossup, leaning Kerry. Electoral-vote.com has to be joking about Hawaii, and NH will not turn its back on its neighboring state's senator. Nevada is hard to read because Las Vegas (which will tend Democratic) is growing faster than the rural (Republican) areas.

If Kerry wins either Ohio or Florida, then Bush must win both Colorado and New Mexico to win. I'm not ready to declare Ohio as this year's one and only pivotal state, but we're kidding ourselves if we don't admit that Ohio and Florida are the important states. Keep in mind that Ohio is in the eastern time zone and Florida has a small part in the central time zone. Thus, to be fair, the networks should hold off on calling Florida (not that they won't be a little gun-shy about calling Florida too early anyway). So in that sense, Ohio is key because we'll know the result sooner. An early Kerry victory in Ohio and all of a sudden FL, MN, IA, WI, CO, and NM become very important. If Bush wins OH and FL, then Kerry needs to sweep CO, NM, IA, and either MN or WI. That's not impossible, but neither is it easy.

No one is going to bed early on Nov. 2nd.

Joe Trippi once again proves that he gets it. I usually resist the temptation to say that this election is like no other. I've studied the elections of the 19th century enough to know that they were just as dirty. The election of 1876 even saw the loser of the popular vote win the electoral vote; that year even had states where the vote was disputed. Yes, Florida was one of them. Point is, the basic issues of presidential politics are not changing nearly as quickly as the mechanics by which those issues filter down to the electorate.

Joe Trippi understands this, and I have to believe it is why the Dean campaign did as well as it did. Whatever your opinion of Dean as a candidate, his campaign reflected a change that is coming. I think this might be the last campaign that a Dean-like candidate loses. So study up, future candidates of both parties. Learn from Dean. It's like I wrote earlier, the Internet is starting to really hit its stride in terms of becoming a vehicle of communication. The Internet makes it possible for meetup.com to exist. Blogs don't replace the regular media, but if they keep the media honest, it's a good thing.

It won't fundamentally change the American people. The issues at stake won't change. But more people will have a voice. I'm not entirely sure that the winner in 2004 (whoever it is) really understands that, but mark my words, the winner in 2008 will.

If you are the type who stays up late on election night, you need to look at this! The electoral map is the best I've seen. Click the "pick 'em" tab to change the map and see who you think will win.

There was some interesting presidential polling data in the news from yesterday. CNN reports that Bush's lead in the Gallup Poll increased to 13 percent (8 percent among registered voters). Meanwhile a poll by the Pew Research Center shows the president's lead going from double digits to no statistically significant lead at all.

When the polls disagree, I always refer back to the Iowa Electronic Market, which has a pretty good record at predicting the popular vote over the years. The IEM is showing roughly a 51 to 49 lead for Bush at the time of this writing. This graph shows how sentiment has changed over the last few months. The popular vote, at any rate, looks too close to call. But as the elections of 2000 and 1888 remind us, it is the electoral vote that counts. Tim Russert said it best in 2000 when he wrote "Florida, Florida, Florida" on his whiteboard before anyone knew how prophetic he would turn out to be.

So is it going to be Florida again? I'm not ready to say yet whether Florida will decide the election, but it, like Missouri, is almost certain to be a bellwether state. A close call in Florida will mean a long night on November 2nd whether they cast the deciding electoral vote or not.